"Computer Communication Theory" Questions - Absolute Wonderful

This article: http://www.why100000.com/_articles/show_a_article.asp? Tab = TabDocument & AutoID = 14 More Content, please visit "100,000 Why" Computer Learning Network: http://www.why100000.com -------------------------------------------------- ----------------------------------------

1-01 The development of computer networks can be divided into several phases? What are the features of each stage? A: The development of computer networks can be divided into the following four phases. (1) Computer communication network for the terminal: It is characterized by the computer is the center and controller of the network. The terminal is distributed around the center computer, which is a hierarchical star structure, and each terminal shares the hardware and software resources of the host through the communication line. The main task of the computer is still batch, and after a time-time system in the 1960s, there is interactive processing and a batch processing capability. (2) Group Switching Network: The packet switched network consists of communication subnets and resource subnets, with communication subnets, not only sharing the resources of communication subnets, but also share the hardware and software resources of resource subnets. The network sharing uses the queuing method, that is, the packet switch of the node is responsible for packet storage forwarding and routing, which gives two user segments (or dynamic) allocated transmission bandwidth, which can greatly improve the utilization of communication lines. It is very suitable for burst computer data. (3) Form a computer network architecture: In order for computer networks of different architectures, international standardization organization ISO proposes a standard framework that enables various computers to interconnected internet networks around the world - open system interconnect basics Model OSI .. In this way, as long as the OSI standard is followed, a system can communicate any other system of the same standard in any place in the world. (4) High-speed computer network: It is characterized by the use of high-speed network technology, the implementation of the integrated business digital network, the rise of multimedia and intelligent networks.

1-02 Exploses the advantages and disadvantages of layered processing methods for network protocols. A: Advantages: (1) allows each other to independent, and a layer can use the service provided by its next layer without knowing how the service is implemented. (2) Good flexibility, when a certain layer changes, as long as its interface relationship is unchanged, the layers above or below are not affected. (3) Structure can be separated, and each layer can be achieved by the most suitable technique. (4) Easy to implement and maintain. (5) Promote standardization. Disadvantages: Hierarchical division is too strict, so that the service provided by the lower layer cannot be called, and the protocol efficiency is reduced.

1-03 Test a method of processing a network layer in a broadcast network. Discuss whether this layer is needed? A: The broadcast network is a shared broadcast channel. There is no routing problem, but can not be a network layer, but from OSI's view, network devices should be connected to the network layer service access point, so set the service access point in high-level protocol and The IEEE 802 standard is processed on the interface of the logical link sublayer in the data link layer.

1-04 Try to compare the architecture of TCP / IP and OSI. Discuss its excessive differences. A: (1) The same point of OSI and TCP / IP is that both are hierarchical, and are all hierarchical. (2) Different points from OSI and TCP / IP: 1,7, from bottom to date, data link layer, network layer, transportation layer, session layer, representation layer, and application layer, and TCP / IP Four layers: Network interface layer, network mesh (IP), transport layer (TCP), and application layer. Strictly speaking, the TCP / IP network protocol only includes the next three floors, the application is not part of TCP / IP. 2 OSI hierarchical relationships, two (n) layer entities must pass through the next layer (N-1) layer entity, can not be cross-level, and TCP / IP can be used directly with the next layer directly using lower levels. The services provided (this level of relationships are often referred to as "grade" relationships), thus reducing some unnecessary overhead and improving the efficiency of the protocol. 3osi only considers the use of a standard public data network to connect a variety of systems together, and later recognizing the importance of the Internet protocol, only a sublayer is drawn from the network layer to complete the interconnection. TCP / IP is considering the interconnection problem of a variety of heterogenes, and Internet protocol IP is an important part of TCP / IP. 4osi started to be biased to the connection-oriented service, and then began to develop unconnected service standards, and TCP / IP has a connection and connectionless service, non-connection service, data transfer, data transfer, and packet voice communication in the Internet It is very convenient. The emphasis on reliability between 5osi and TCP / IP is also different. Both connection services, data link layers, network layers, and transportation layers, especially in data link layers, providing reliability, and network and transportation layers, especially in data link layers with verification, confirmation and timeout retransmission. There are also similar techniques. TCP / IP, TCP / IP believes that reliability is an end-to-end problem, which should be resolved by the transportation layer, so it allows a single link or machine to lose data or data error, the network itself does not perform error recovery, loss or The recovery of the error data is performed between the source host and the destination host, and is completed by the transportation layer. Since reliability is completed by the host, the host's burden is added. However, when the application is not high, even the host does not have to perform reliability processing, in this case, the TCP / IP network has the highest efficiency. 6 The location in the two architectures is also different. The OSI network layer provides a connection-oriented service, which is incorporated into the network service, and is incorporated into the network service, and there is not much thing to leave the end host. Instead, TCP / IP requires hosts to participate in almost all network services, so the host requirements for entering the network are high. 7osi starts to consider network management issues, and then consider this problem, and TCP / IP has better network management. 1-05 Which aspects of computer network can be classified? A: Computer network can be classified from different angles: (1) divided into circuit exchange, packet exchange, packet exchange, and mixing exchange according to network switching function; (2) can be divided into star network according to the topology of the network. Tree network, bus network, ring network, mesh, etc. (3) according to network communication performance can be divided into resources sharing computer network, distributed computer network, and remote communication network; (4) According to the network coverage Scale can be divided into local area network, metropolitan network and wide area network; (5) divided into public networks and private networks based on the network of use.

1-06 Which part of the computer network consists of? A: A computer network should have three main components: (1) Several hosts, providing services to each user; (2) a communication subnet, it is made by some dedicated node switches and connecting these nodes The link is composed; (3) a series of protocols, which are used for communication between the host or between the host and the subnet. 1-07 The Red Army of the blue army occupied by the two hills and the Red Army of the valley stationed between the two mountains. The comparison of its strength is: a blue army on the top of a mountain, but the blue army in the top of the mountains can fight the Red Army. The blue army on the top of the mountain is planned to attack the Red Army on the next day. So send a call to another mountain. However, communication lines are very bad, and the possibility of errors or loss is large. Therefore, a friendly army that requests the call must be sent back to a confirmation. But this confirmation can also be wrong or lost. Can I design a protocol that makes the Blue Army to achieve a collaborative combat, which is a certain (ie 100%) to win? A: It is impossible to design such an agreement. The last transferred message also needs to confirm, then which message is the last message? Note: It is required to be 100% reliable, not 99.9999 ...% reliable.

1-08 What is the features of the connection services and connectionless services? A: The connection service must be established before data exchange, keep the underlying resource, after the data exchange is over, the connection should be terminated, and the reserved resources are released. For connectionless services, no connection between the two entities can communicate, dynamically allocate the underlying resource during data transmission, and do not need pre-reserves in advance.

1-09 What is the difference between the agreement and service? What is the relationship? A: The protocol is horizontal, and the service is vertical.

1-10 What is a computer network? What consists of computer networks? A: (1) Computer network is to share information sharing and capabilities between computers through communication tools. (2) Computer network consists of computer subnets and communication subnets.

1-11 What is the difference between computer multi-user systems and network systems? A: Computer network is to share information sharing and capabilities between computers. Computer network features network communication, resource management, network services, network management, and interactive ability. Network implementation information exchange and resource sharing distributed in different geographic locations, computer resources are mainly computer hardware, software and data, data is a carrier of information, and network users can use local computer resources while using network access. The resources on the remote networking computer can even call multiple computers in the network to complete a task. However, multi-user systems refer to software systems, which can set access to users, there is conditionally to use the same computer resources to use the same computer resources, and can be used for information exchange and resource sharing.

1-12 What is the topology map of a computer network? A: The topology map of the computer network refers to the geometric arrangement of using topologically connects the communication lines and sites (computers or devices) in the network.

1-13 Communication Subnet and resource subnets consisting of major parties? What is the main function? A: The communication subnet consists of two parts of the transmission line and the switching unit. The main function is to transfer the message from one host to another. The resource subnet is a collection of various network resources, and its main function is the computer information exchange and resource sharing.

1-14 What kind of type of computer network is divided into computer network? Try to compare the characteristics of different types of networks. A: 1. Classification by network coverage: (1) LAN: LAN is a network composed of computer hardware in a relatively small range of communication lines, generally limited to a smaller area, usually connected in a wired manner. (2) Metropolitan area network size is limited to a city, covered by dozens of kilometers to hundreds of kilometers, the metropolitan area network is basically the extension of the local area network, which is usually used similar to the local area network, but The range is relatively wide in transmission media and wiring structures. (3) WAN: The coverage is very wide, also known as the remote network, which is different from the technology, application scope and agreement standards. 2, according to the hemage medium: (1) Wired network: a computer network connected to the coaxial cable, twisted pair, and even using a wire TV cable cable, and the wire network transmits information through the "carrier" space, it needs to be used Wire is implemented. (2) Wireless network: Transfer media with air, use electromagnetic wave as a carrier to propagate data. Wireless network includes: wireless telephone, voice broadcasting network, wireless TV, microwave communication network, satellite communication network. 3, classify according to the topology of the network: (1) Star network: Each site is connected to the center by point-to-point, which is easy to add new sites to the network, data security and priority easy control It is easy to realize network monitoring, but once the central node is faulty, it will cause the entire network. (2) Bus type network: All sites in the network share a data channel, the bus type network is installed simple and convenient, the shortest wire, low cost, a site failure generally does not affect the entire network, but the media failure will Leading the network paralysis, the bus network is low, and the monitoring is difficult, and the new site is not as easy as the star network. (3) Tree network: It is the synthesis of the above two networks. (4) Ring network: The ring network is easy to install and monitor, but the capacity is limited, after the completion of the network, adding new sites more difficult. (5) Mesh network: The mesh network is a comprehensive application based on the above topological networks. 4. Classify: (1) Point-to-point transmission network: Data is transmitted in a computer or communication device in a point-to-point manner, and there is many ways to connect between a pair of machines. the way. (2) Broadcast transmission network: Data is transmitted in a shared communication medium line, shared a communication channel from all machines on the network, which is applicable to a network of small networks or non-confidential requirements for small networks or secret requirements.

1-15 What is a network architecture? Why define the network architecture? A: The architecture of the computer network is to provide corresponding specifications and standards for interconnect and interoperability between different computers. First, data transmission issues must be resolved, including data transmission mode, errors in data transfer, and errors, transmission network resource management, communication addresses, and file formats. Solving these issues between computers between intercom communications and frequent consultations between computers and communication networks. These negotiations and adjustments and information transmission and reception can be designed and implemented in different ways. The most important framework file in the computer network architecture is a computer network 7-layer open system interconnect standard for international standardized organization. Its core content contains high, medium, low-top three layers, high-level face-to-network applications, low-level physical devices, and intermediate layers, information conversion, information exchange (or transfer), and transmission path selection, etc. That is, the routing selects core. Computer network is a very complex system. It integrates contemporary computer technology and communication technologies, but also related to knowledge and technology in other applications. The software and hardware system, different communication networks, and various external auxiliary equipment are connected to the network system, and the information sharing in the high-speed and reliable information sharing is the main problem facing computer network. In order to solve this problem, people must define a network system Different computers, different communication systems and different applications can connect (interconnect) and interconnect (interoperability) open network architectures. Interconnections means that different computers can connect to each other through communication subnets to communicate with each other. Interoperability means that different users can use the resources and information in other computers on the connected computer, using the same command or the same operation, as using the resources in the local computer system. 1-16 What is a network protocol? What is its role in the network? A: The rules, standards or agreements established in the network are called the network protocol. Mainly by grammar, semantics, and synchronization (referring to the order in the event implementation). Communication protocols have hierarchical features, most of the network organizations organize in a layer or level, based on the next layer, and the purpose of each layer is to provide a certain service to it. Show the details of this service to the previous layer. The network protocol determines the exchange data format and the relevant synchronization problem.

1-17 What is the 0si reference model? What is the main function of each layer? A: The most important framework file in the computer network architecture is the 3-layer open system interconnection standard for the international standardization organization, namely the 0SI reference model. Its core content contains high, medium, low-top three layers, high-level face-to-network applications, low-level physical devices, and intermediate layers, information conversion, information exchange (or transfer), and transmission path selection, etc. That is, the routing selects core. This model proposes the interconnection and interconnection function of the computer network with a layered method. The layering is to divide a complex problem into different local problems and stipulate the functions that must be completed. The layering decomposes the complex problem into a plurality of relatively simple issues, and enables high-level users to interconnect from the protocol layer having the same function. This makes the system become open. Physical layer: Responsible for providing and maintaining physical lines, and detects processing contention conflicts, providing end-to-end error recovery and stream control transmission in bits. Data Link Layer: The main task is to enhance the function of the physical transmission raw bits and transmit in units of frames. Network layer: It is related to the running control of the subnet, one of which is to determine the "routing" of the packet from the source to the destination to transmit in units. Transportation layer: The basic function is to receive data from the session layer, and it is necessary to divide it into a smaller unit transfer, and ensure that the information reaches the other party is correct. The transportation layer must also decide to serve the session layer and eventually serve the network user. The session layer: performs high-level communication control, allowing users on different machines to establish session relationships. Indicates the layer: Complete certain features. For example: solving the conversion of data formats. Indicates that the layer is concerned about the syntax and semantics transmitted, and the layers below the layer are only concernedically transmitted and the bitstream is reliably transmitted. Application Layer: Provide information related to user application, including web browsing, email, file transfer, virtual terminal software, process job input, directory query, and other versatile functions, etc. 2-01 What problems should I solve the physical layer? What is the main feature of the physical layer? A: (1) The main problem to be solved by the physical layer: 1 The physical layer should be different as possible to block the physics equipment, transport the media and communication means, so that the above data link layer does not feel these differences, but focused on Complete the protocol and service of this layer. 2 The ability to serve and receive bitstreams (typically serial transmissions transmitted sequential transmissions) on a physical transmission medium. To this end, the physical layer should address the establishment, maintenance and release of physical connections. 3 Uniquely identify the data circuit between the two adjacent systems. (2) Main features of the physical layer: 1 Since many physical procedures or protocols have been developed before OSI, these physical procedures have been adopted by many commercial devices in the field of data communications. In addition, the physical layer protocol has a wide range, so it is not based on the abstract model of OSI to develop a new physical layer protocol, but the existing physical procedures, the physical layer is determined to describe the mechanical, electrical, electrical, electrical, and transmitting media interfaces. , Function and procedure characteristics. 2 Due to many ways of physical connection, there are many types of transmission media, so specific physical protocols are quite complicated.

2-02 What are the commonly used transport media? What are the features? A: The common transmission media has the following: (1) Twisted pair: score shield twisted pair and no masked twisted pair. Composed of two mutually insulated wires. The analog signal can be transmitted, or the digital signal can also be transmitted, and the effective band is 250 kHz, and the communication distance is generally several to ten kilometers. The farther the wire is, the farther the communication distance. When digital transmission, if the transmission rate is several megabits per second, the transmission distance can reach a few kilometers. It is generally used as telephone line transmission sound signal. Although the twisted pair is easily interfered with external high-frequency electromagnetic waves, the error rate is high, but because of its cheap, it is easy to install, it is suitable for point-to-point connection, but also for multi-point connection, so it is still widely used. (2) Coaxial cable: Sub-band coaxial cable and broadband coaxial cable, its structure is in addition to a layer of insulated solid wire, and then there is a layer of insulated hollow circular wires in the outer layer. Due to its high bandwidth (up to 300 ~ 400 Hz), low error rate, performance price ratio, in LAN, the maximum transmission distance of coaxial cables is different from cable model and transmission signal, due to susceptible low frequency interference Module the signal in the high frequency carrier when used. (3) Optical fiber: The optical signal is propagated as a carrier as a carrier. The advantage is that the diameter is small, the weight is light; the transmission frequency bandwidth is high, the communication capacity is large; the lightning and electromagnetic interference performance is good, no stroke interference, good confidentiality, and the error rate is low. However, the price of photoelectric interfaces is more expensive. Optical fibers are widely used in the telecommunications system layout. (4) Wireless channel: Separate microwave relay communication and satellite communication. Its main advantages are that the frequency is high, the bandwidth, and the communication channel is large; the signal is smaller, the transmission quality is high, and the communication is relatively stable; the influence of the geographic environment, the construction investment is small, fast. Disadvantages is that the ground microwave relay communication is spread in space, and the transmission distance is limited, generally only 50km, concealed, and confidentiality. Satellite communication although the communication distance is far and the communication cost is independent of the communication distance, the spread is delayed, the technology is more complicated, and the price is more expensive. 2-03 What is the characteristics of the baseband signal and the broadband signal? Answer: (1) Use the digital signal "1" or "0" to represent two different voltages, which constantly alternately alternate signals, and baseband is the original signal. Basic band. The baseband signal is delivered directly to the line transmission is called a baseband transmission. The baseband transmission requires a wide frequency band. (2) If the multi-channel baseband signal, the spectrum of the audio signal, and the video signal is transferred to different frequency bands of a cable, such a transmission method is called broadband transmission. The signals transmitted by the broadband transmission are modulated analog signals. It is therefore possible to achieve integrated transmission of text, sound, and image with a broadband transmission system. In the broadband system, you need to increase the transmission distance with the amplifier.

2-04 There is 600MB (megabytes) data, you need to transfer from Nanjing to Beijing. One way is to write the data to the disk, and the trustee will take these disks by train. Another method is to transmit this data with a computer through a long-distance telephone line (setting information transfer rate is 2.4 k / s). Try comparing the advantages and disadvantages of these two methods. If the information transfer rate is 33.6kb / s, what is the result? A: Assume that continuous transmission is not wrong. It takes 24.3 days if used in 2.4 kb / s (= 600 × 1048576 × 8 = 5033164800 bit). If you transfer with 33.6kb / s, you need to 1.73 days. Bidders are slow, and more expensive. 2-05 56kb / s modem breaks through the path limit transmission rate of Shannon? What is the condition of this modem? Answer: 56kb / s modem is mainly used for communication of users and ISP. At this time, only one A / D conversion is only once between users to ISP, and the quantitative noise of the 33.6kb / s modem used between the two users. Small, so the signal to noise ratio is further improved. Although the rate of 33.6kb / s modem has basically reached the path limit transmission rate of Shannon, the use conditions of the 56kb / s modem have different conditions, it improves the signal-to-noise ratio, which does not break through the formula of the permanent transmission rate. The use of 56kb / s modem is that ISP also uses such a modem (here to perform conversion between digital signals, rather than digital-to-mode conversion), and is a digital channel between ISP and telephone switches. If the ISP is just a 33.6kb / s modem, the user-end 56kb / s modem automatically reduces the same rate as the 33.6kb / s modem.

2-06 When introducing the twisted pair, we said: "When digital transmission, if the transmission rate is several megabits per second, the transmission distance can reach a few kilometers." But now we use the modem to connect to ISP, data The maximum transmission rate can only reach 56kb / s, far from several megabits per second. Why is that? A: "When digital transmission, if the transmission rate is several megabits per second, the transmission distance can reach a few kilometers." This refers to the use of digital lines, and its two ends have no bandwidth limit. When we use the modem to connect with the ISP, use the user line of the phone. Such a user line enters the local call to limit the bandwidth of 3400 Hz and the bandwidth of the digital line is very different.

2-07 Test comparison of the advantages and disadvantages of simulated communication methods and digital communication methods. A: In the data communication technology, the method of transmitting analog data signals by modem is transmitted by modem, called broadband transmission. The method of directly transmitting digital data signals using a digital communication channel is called baseband transmission. The advantage of broadband transmission is that an analog voice communication channel that is the most widely covered, universal application. Telephone exchange network for voice communication is mature, low cost, its disadvantage is that the data transmission rate is low, and the system efficiency is low. Baseband transmission directly transmits digital signals without changing digital data signal bands (ie waveform), it can achieve high transmission rates, which is currently actively developing and widely used data communication.

2-08 Describe how the bridge is used to reduce network traffic problems?

A: The bridge operates in a "mixed" manner, which means that it is to view its address when sending each package, and the bridge works in the Media Access Control (MAC) sub-layer of the 0SI data link layer. The bridge intercepts all network "traffic" and checks each package it receives, reads out the destination address of each package, thus determines whether the package forwards the package to the next network. If the goal of the package is a local node, the bridge filtering the package on the source LAN. Therefore, the bridge can split the traffic between the two networks, which facilitates improving the performance of the interconnect network. 2-09 What network conditions can cause network segment being separated from partitioning? A: When there is a problem in the network segment, and the network segment contains a lot of traffic, it is isolated from the network.

2-10 What is a multiplexer? A: In the channel multiplexing technology, the data is combined from multiple input lines, and the data signal is sent to a high-capacity data link, so that different computers are connected to the same channel, share The device of the channel resource is a multiplexer.

What is the difference between 2-11 bit rate and baud rate? A: Baud Rate: The number of discrete signal events per second. Bit rate: number of bits per second.

2-12 In the case of how to install multi-port repeaters? A: The network repeater enlays the input signal and re-generates the signal along a plurality of running cables. Once the signal is reauthenticated, it helps to avoid conflicts. When the multi-port repeater receives the packet, it is reordered, and the package can be placed on the extended network segment, if the repeater detects a problem, for example: excess conflict, it stops the transfer will stop transfer data To the cable segment. A network segment on the multi-port repeater is separated without affecting other ports. Once the network problem is resolved, the spaced network can be reset on the repeater to restore the transfer. Therefore, multi-port repeaters can be employed when connecting multiple users that are derived from the distance.

2-13 Compare the nature and characteristics of different transmission media. A: 1. Twisted pair: physical properties: twisted pair consists of 2 or 4 insulation lines arranged in rule spiral structures. A pair of lines can be used as a communication circuit, and the purpose of each line to the helix is ​​to minimize electromagnetic interference between each line pair. Transmission characteristics: The most common application of twisted pair is analog transmission of speech signals. When using a twisted pair to transmit analog data signal by a modem, the data transmission rate is currently up to 56 kb / s, and the two-way 33.6kb / s, and the total data transmission rate of 24 audio channels can reach 230kb / s. Digital data signals are transmitted using twisted pairs, and the general data transmission rate is up to 2MB / s. Connectivity: Twisted pairs can be used for point-to-point connections, and can also be used for multi-point connections. Geographical range: When the twisted pair is used for a remote trunk, the maximum distance can be up to 15 kilometers; when used for 10 Mb / s LAN, the distance to the hub is up to 100 meters. Anti-interference: When the low frequency transmission, its anti-interference ability corresponds to the coaxial cable. At 10 --- 100 kHz, its anti-interference ability is lower than the coaxial cable. Price: The price of twisted pair is lower than other transmission media, and is installed and maintained. 2, coaxial cable: physical characteristics: coaxial cables are also composed of two conductors, with thickness, which is composed of hollow cylinders in a single-in-cell conductor. The inner conductor is a solid or twisting; the outer conductor is a whole or textile. The inner conductor is fixed with an insulating ring or a hard electric vehicle of the rule pitch, and the outer conductor is packaged with a sheath or shield. Transmission features: 50 ohms are dedicated to digital transmission, generally using Manchester coding, with data rate up to 2MB / s. The CATV cable can be used to simulate and digital signals. Frequency of the analog signal, up to 300-400 MHz is possible. For digital signals, it has reached 50Mb / s. Connectivity: Coaxial cable can be used for point-to-point connections, or for multi-point connection. Geographical range: The maximum distance of the typical baseband cable is limited to several kilometers, and the broadband network can extend to the range of tens of kilometers. Anti-interference: The structure of the coaxial cable makes it strong, and the anti-interference resistance of the coaxial cable depends on the application and implementation. Typically, for higher frequencies, it is due to the anti-interference of twisted pairs. Price: The cost of the coaxial cable with good installation quality is between twisted pairs and optical fibers, and is convenient for maintenance. 3, fiber optic: physical properties: optical fibers are a medium of diameter fine (2 --- 125 microns), which can conduct a medium that can conduct light waves, and can conduct a medium. Various glass and plastic can be used to make optical fibers. The optical cable has a cylindrical shape, consisting of three concentric parts: core, cladding, sheath. Transmission characteristics: Optical fiber utilizes full internal reflection to transmit signal encoded beams. Different mode and single mode, multimode bandwidth is 200 MHz --- 3GHz / km; single mode has a bandwidth of 3 GHz - 50GHz / km. Connectivity: The most common use of fibers is on the link of point-to-point. Geographical range: The fiber signal is extremely small, which can do not use the repeater to achieve high rate data transfer within 6 --- 8 kilometers. Anti-interference: is not affected by electromagnetic interference and noise disturbance. Price: The current fiber system is more expensive than the twisted pair system and the coaxial cable system, but with the advancement of technology, its price will fall to compete with other media.

3-01 What is the difference between data link (ie, logical links) and links (physical links)? "Circuit is connected" with "Data Link"? Answer: (1) The difference between the data link and the link is that the data link is removed, and there must be some necessary procedures to control the transmission of data. Therefore, the data link is more than the hardware and software required to implement the communication procedure than the link. (2) "The circuit is connected" indicates that the node switch across the link has been turned on, and the physical connection has been able to transmit bitstream. However, data transmission is not reliable. On the basis of physical connection, establish a data link connection is the "data link is connected". Thereafter, since the data link connection has functions such as detection, confirmation, and retransmission, the less reliable physical link becomes a reliable data link to perform reliable data transmission. When the data link is disconnected, the physical circuit connection does not necessarily follow the disconnection. 3-02 Testing stop waiting agreement algorithm. At the receiving node, when the step (4) is tried, if it is "otherwise to (7)" change to "otherwise to (8)", what results will result? A: "Otherwise" means that the N (S) of the frame transmitted by the sender is different from the status variable V (r) of the receiver, indicating that the sender does not receive the ACK sent by the receiver, so retransmit the last frame. If "Go to (8)", the recipient is to send NAK, and the sender continues to retransmit the last frame, which is always like this.

3-03 The channel rate is 4kb / s. Use stop waiting agreement. Communication TP = 20ms. Confirm that the frame length and processing time can be ignored. How much can I ask the frame length to reach at least 50% of the channel utilization? A: Due to ignore the bit rate, confirm the frame length, and processing time, the transmission time relationship of the response frame of the data frame in the protocol is stopped. Assume that the frame length is L, the channel rate (ie, the data transfer rate) is S. The channel utilization is based on the meaning of the question, so that is, that is, the Bit, the frame length is more than or equal to 160% to make the channel util of at least 50%. .

3-04 Try to briefly describe the meaning of each field of HDLC frames. What method is used to ensure the transparent transmission of data? Answer: (1) The format of the HDLC frame, the information field (length variable) is data of the data link layer, which is a packet from the network layer. At both ends of the information field are 24 bit of the frame header and the frame end. The flag fields at both ends of the HDLC frame are used to define the boundary of a frame, and the address field is used to fill in the address information of the slave station or the answer station, and the frame check sequence FCS is used to verify the bit stream composed of the address, control, and information fields. The control field is the most complex and used to implement many main functions. (2) The transparent transmission of the link layer is achieved by zero bit filling method, that is, six consecutive 1 between the two flag fields. The specific approach is to send the end when a binder is not added to the flag field, and then discovers 5 consecutive 1, then insert 1 0, and find f first in the receiving end The field is determined the boundary of the frame, and then the bit stream therein is scanned, and each of the five consecutive 1 after 5 consecutive 1 is found to be reduced to the original bitstream.

3-05 Data Link Agreement Almost always puts the CRC on the tail instead of putting on the head, why? A: CRC is calculated during the transmission. Once the last bit is sent to the outgoing line, the CRC encoding is immediately added after the output stream. If the CRC is placed in the header of the frame, then the entire frame first checks the CRC before sending. This should be handled twice, and the first time is to calculate the check code, the second time is to send it. Put the CRC on the tail to halve the processing time. 4-01 Try comparing the characteristics of several shared channels. A: The shared broadcast channel uses channel-based sharing and two ways based on queuing. Channel sharing can be used in frequency division multiplexing or time division multiplexing, no matter which technique is used, there is a fixed allocation and two different ways. Based on the queuing sharing can be assigned bandwidth in two ways: one is random access, that is, allowing the station to freely send data. When a conflict occurs, the conflict is resolved by a certain algorithm. Another method is to try to form a distributed logical queue or use token to coordinate the sends data from each station. These four ways to share broadcast channels: fixed allocation methods are good in real time, but the channel utilization rate is low; the channel utilization of the assignment method is high, but the workstation must increase certain processing power, and the channel is busy, some users have channel The application may be blocked, and the application is delayed; the random access method is simple, the workstation access is convenient, when the network is low, the network is substantially no delay, but the transmission is uncertain, the network is Efficiency has dropped a lot; distributed logical queue or token method, time delay determination, can set priority, can transmit digital packet voice signals, the heavy load performance, but the protocol is complicated.

4-02 In the pure Aloha protocol, if the system works in the state of g = 0.5, the credit is the probability of idle. A: For pure Aloha, the probability of the K frame is generated within any frame from the Poisson distribution. The channel is idle (ie: generating 0 frame) The probability of E. -.5 = E of -0.5 times = 0.6055.

4-03 10,000 terminals to compete with a common time slot Aloha channel. Each terminal transmits frame 18 times per hour, the slot length is 125Us, and the channel load G is tried. A: Each terminal transmits 1 frame per 200 (= 3600/18), with a total of 10,000 terminals, therefore, the total load is 100,000 frames in 200 seconds, and the average 50 frames per second. 8,000 time slots per second, the average number of times each time slot is g = 50 ÷ 8000 = 1/160.

4-08 There are 4 stations for code multiplexing multiplex. The 4-station chip sequence is: A: (- 1 -1 -1 1 1 -1 1 1) B: (- 1 -1 1 1 1 1 1) C: (- 1 1 -1 1 1 1 -1 -1) D: (- 1 1 -1 -1 -1 -1 1 -1) Received such a chip sequence: (- 1 1 -3 1 -1 -3 1 1). Which station is sent? What is the code sent? A: Only 4 conventional metals: (-1 1 -3 1 -1 -1 -3 1 1) · (-1 -1 -1 1 1 -1 1 1) / 8 = 1 (-1 1 - 3 1 -1 -3 1 1) · (-1 -1 1 -1 1 1 1 -1) / 8 = -1 (-1 1 -3 1 -1 -3 1 1) · (-1 1 -1 1 1 1 -1 -1) / 8 = 0 (-1 1 -3 1 -1 -3 1 1) · (-1 1 -1 -1 -1 -1 1 -1) / 8 = 1 result is a And D transmission bit 1, b transmission bit 0, and C does not send data. 4-09 Main features of the LAN? Why is the local area network a communication network?

A: LAN LAN refers to a computer communication network that connects limited communication devices within a smaller geographic range. From a functional point of view, the LAN has the following features: 1 Sharing transport channel. In a local area network, multiple systems are connected to a shared communication medium. 2 Limited geographic range, limited users. Usually the local area network is only one unit service, only in a relatively independent local area, such as a building or a concentrated building group. In general, the coverage of the local area network is approximately 10m to 10km or more. 3 High transmission rate. The data transmission rate of the local area network is generally 1 to 100 Mbps, which can support high-speed communication between the computer, so the delay is low. 4 The error rate is low. Due to close transmission, the error rate is very low, generally between 10-8 ~ 10-11. 5 distributed control and broadcast communication. Each station in the local area network is equally related to the main relationship, broadcast or multicast can be broadcast. From the network architecture and transmission control procedures, the LAN also has its own characteristics: 1 Low-level agreement is simple. In the local area network, due to the short distance, low cost, low cost, high transmission rate, high reliability, the channel utilization is not the main factor considered, so the low-level protocol is relatively simple. 2 The network layer is not set separately. The topology of the local area network uses shared channels such as bus type, ring type and star, and the network generally does not require intermediate transfer, flow control and routing function is greatly simplified, and network layers are typically established in the LAN. Therefore, the architecture of the local area network is only equivalent to the least two layers of OSI / RM. 3 Use a variety of media access control techniques. Due to the use of shared broadcast channels, the channels of the local area network are used by different transmission media, the problem of the local area is a multi-source, multi-purpose link management. This leads to a variety of media access control technology. In the architecture of the OSI, a communication subnet has only the lowest three layers. Only the system structure of the local area network is only the lower three floors of OSI, there is no level of the fourth layer. Therefore, the local area network is just a communication network.

4-10 What is the difference between the IEEE 802 LAN Reference Model and the OSI Reference Model? A: The architecture of the LAN is very different from the architecture of OSI. Its architecture only has the lower three layers of OSI, without the level of the fourth layer. Even the next three floors, because the local area network is shared broadcast channel, and there are many kinds of products, involving various media access methods, so there is a significant difference. In the local area network, the physical layer is responsible for physical connection and transmitting bitstreams on the media, which main tasks are some of the features of the transmission media interface. This is the same as the physical layer of the OSI reference model. However, due to the variety of transport media, various media can be used in a variety of transport media, so the processing of the physical layer in the local area is more complicated. Typically, the physical layer of most local domain networks is divided into two sub-layers: a sub-layer describes the physical characteristics associated with the transmission medium, and the other sub-layer describes the physical characteristics independent of the transmission medium. In the local area network, the main role of the data link layer is to achieve reliable data transmission on a less reliable transmission channel through some data link layer protocols, responsible for the transmission and control of the frame. This is the same as the data link layer of the OSI reference model. However, in the LAN, due to the sharing of network public channels in each station, it must be resolved how channel allocation, how to avoid or resolve channel contention, that is, the data link layer must have a media access control function. There is a variety of topology and transmission media used in the local area network. There are also many corresponding media access control methods, so it should be separated from the partial and unrelated part of the transmission media in the data link function. Thus, the data link layer in the IEEE 802 local area network reference model is divided into two sub-layers: media access control MAC sub-layer and logical link control LLC sub-layers. There is no network layer in the IEEE 802 LAN Reference Model. This is because the topology of the local area network is very simple, and each site shares the transmission channel. Only one link between any two nodes, does not need to perform routing and traffic control, so the network layer is not set separately in the local area network. . This is different from the OSI reference model. However, from OSI's point of view, the network device should be connected to the network layer service access point SAP. Therefore, although the network layer is not provided in the local area network, the service access point SAP of the network layer is set on the interface of the LLC sublay and the high-level protocol. From the above analysis, the reference model of the local area network is only equivalent to the least two layers of the OSI reference model, and there is also a big difference between the physical layer and the data link layer of the two. The physical layer and the media access control MAC sub-layer of each sub-standard in the IEEE 802 series are different, and the logical link controls the LLC sub-layer is the same, that is, the LLC sub-layer is actually high-level protocols and any one. Standard interface between MAC sub-layers. 4-12 How much is the symbol transmission rate of Ethernet of 10MB / s? A: The symbol transmission rate is the baud rate. Ethernet uses Manchester code, which means that each bit of sending has two signal cycles. The standard Ethernet's data rate is 10Mb / s, so the baud rate is twice the data rate, namely 20M baud.

4-13 It is assumed that the 1km long CSMA / CD network has a data rate of 1GB / s. The propagation rate set on the network is 200000km / s. Seeking the shortest frame length of this protocol. A: For 1km cable, one-way propagation time is 1 ÷ 200,000 = 5 ×, that is, 5us, and the time to return to 10Us. In order to be able to work according to CSMA / CD, the launch time of the minimum frame cannot be less than 10Us. Working at 1GB / s, the number of bits that can be transmitted by 10US is equal to: 10x10 -6 -6 -9 = 10000, and therefore, the shortest frame is 10,000 bits or 1250 bytes long.

4-14 What is a local area network? What is the character? A: 1. The LAN is a communication system that allows many individuals that are independent of each other to communicate directly in the appropriate area to communicate directly with appropriate transmission rates. 2, the characteristics of the LAN: (1), covering a small geographic range, about a few kilometers of geographic scope, is a unit owned, the geographical range and the number of sites are limited, all sites share higher total total bandwidth, ie Higher data transmission rate; (2), local area network is a communication network, with a smaller delay and lower bit error rate, data (bits) transmitted from one connected device to another Range; (3), formation of equal relationships between the sites rather than the primary relationship; (4), can broadcast or multicast (also known as multicast). 4-15 Brief Description Common IEEE802.3 and IEEE802.5 LAN protocol architecture. Answer: IEEE802.3: This standard defines the network developed by Ethernet, and LLC and Mac (media access control sub-layer) of the data link layer, which completes many of the network layer, mainly responsible for "error" actual transmission The channel transformation is a reliable transmission channel, has a medium access control function, and provides a plurality of media access control methods. The MAC sublayer uses a competitive access technology called carrier-listening multiplexing / collision detection (CSMA / CD). This technique uses the network to determine if it is idle to reduce the impact range of conflicts, and attempt to pass the data can only be passed when the network is idle. This reduces conflicts, but does not eliminate conflicts, because the signal is propagated in the network that requires time, but also continues to listen, so it can detect conflicts will occur. When the conflict occurs, all the equipment stops transmitting, and issues a "quot; congestion" signal to inform all conflict sites. Each device needs to wait for a while before retransmission. These security measures are used, obviously Reduces the network conflict, but for the busiest network is not so effective. Ethernet architecture is based on CSMA / CD access method. IEEE802.5: This standard defines the transfer structure of token using tokens, and data link The LLC and Mac (Media Access Control Subside) of the road layer, complete a lot of functions of the network layer, mainly responsible for transforming the actual transmission channel of "errors" into a reliable transmission channel, has a media access control function, and provides media Access control method. MAC uses token frame access technology, the physical topology of the token network is a ring type, use the logical ring station to pass the token, each node must be connected to a hub, it is called multiple access unit MAU The token network is connected to the truncane coupler, and the trunk coupler is also known as a repeater, there are two ways to send and listen, and each site is not in the state of sending data, it is in listening state. Token In fact, a special frame is usually not flowing on the loop. When a station is required to send, it is necessary to intercept the loop, and once the trunk coupler has a token in the bitstream of the loop input, First, the unique flag of the token is first transformed into a frame of the frame (ie, is called interception), followed by placing the trunk coupler of the station to the transmission mode, and transmits the data of the transmit buffer from the loop output end of the trunk coupler. Going out. The architecture of the token network is based on the token access method.

4-16 What is the difference between the IEEE802 LAN Reference Model and 0Si Reference Model? Why is the IEEE 802 standard have been widely used? A: The 0Si architecture refers to the standard reference model of the 7-layer open interconnection standard, the 0Si model is based on the recommendations of the International Standardization Organization, which is the first step in the use of international standardization agreements. This model is called IS0 0SI open system interconnect reference model, which is about how to connect mutually open systems. It is important to note that the 0SI model itself is not all content of the network architecture, because it does not describe protocols and services for each layer, it only shows what should be done each layer. IS0 has set standards for each layer, but they are not part of the reference model, but as an independent international standard. IEEE802 is a series of standard documents issued by the International Electronics and Electrical Engineers Association on the structure of office automation and light industry LAN, which basically corresponds to the physical layer and data link layer of the 0Si model, this standard makes the network physical connection and access. The method is standardized and has been received by IS0 as a standard. Therefore, the IEEE 802 standard has been widely used. 4-17 Briefly describe the two LAN working principles of Ethernet and Token-Ring. A: The Ethernet MAC sub-layer uses a competitive access technology called carrier-listening multiplexing / collision detection (CSMA / CD). This technique uses the network to determine if it is idle to reduce the impact range of conflicts, and attempt to pass the data can only be passed when the network is idle. This reduces conflicts, but does not eliminate conflicts, because the signal is propagated in the network that requires time, but also continues to listen, so it can detect conflicts will occur. When the conflict occurs, all devices stop transmitting and issue a "congestion" signal to inform all conflict sites. Each device needs to wait for a while before returning. The combination of these security measures significantly reduces the network conflict, but it is not so effective for the busiest network. The Ethernet architecture is based on a CSMA / CD access method. The MAC sub-layer of the token network uses token frame access technology, the physical topology of the token network is a loop type, and the use logic ring will pass the token. Each node must be connected to a hub, which is called multiple access. Unit MAU. Each station of the token network is connected to the trunk coupler, and the trunk coupler is also referred to as a forwarder, there are two ways to send and listen, and each site is not in the state of transmitting data, in the listening state. The token is actually a special frame, usually constantly flowing on the loop, when a station is required to send, first intercept the loop, once the trunk coupler appears to make a token in the bitstream of the loop input When the token is first transformed into a frame of the frame (ie, it is called interception), then the truncation coupler of the station is set to the transmission mode, and the data of the transmit buffer is output from the loop coupler. The end is sent. The architecture of the token network is based on token access methods.

4-18 What is the advantage of using twisted pairs than coaxial cables? A: The price of twisted pair is lower than the coaxial cable, and is installed and easy to maintain.

4-19 What is the right network? How to connect?

A: Each computer is equally, which allows resources within other computers. This network is called peer local area network, referred to as peer network. Connection method: First install the same interface network card, connect each computer through the network cable and HUB, after installing, start the computer, install the NIC driver, and "Control Panel / Network" in Windows 95/98 Install "IPX / SPX Compatible" Quot; and "Netbeui Protocol", and click "File and Print Share" button, select "Allow other users to access my file" and "Allow other computers to use my printer" two options, Select the resource that needs to share and the login method, you can realize the main points of the peer-to-peer. A: In the packet switched network, the data is transmitted in a short packet. Typical packet length The upper limit is 1000 bytes (or eight-bit groups). If a source station has a long message to send, the message will be split into a series of packets. Each group contains a part of the user data (or one The short message is all added to some control information. The control information should at least include the information required to route the packet to the destination. Each node on the path is received, a short time storage , Then pass it to the next node. The main advantages of the packet switched network: 1 Efficient. Dynamically allocate the transmission bandwidth during group transmission, the communication link is a tape-segment. 2 Flexible. Each node has intelligence, Separately select forward routes to each packet. 3 rapidly. As a packet as a transfer unit, you can send a packet before communication; the network uses high-speed links. 4 Reliable. Weekly network protocol; distributed multi-route Communication subnet.

5-04 Try to compare the main advantages and disadvantages of circuit exchange, packet exchange and packet exchange from multiple aspects. Answer: (1) Circuit Exchange: Because the circuit exchanges must establish a physical pathway exclusive between the communication before communicating (successful in the exchange equipment between the communication between the communication), therefore is the following Advantage. Advantages: 1 Since the communication line is dedicated to the communication between the communication, the data is straight, the time delay of transmitting data is very small. 2 The physical path between the communication between the communication is established, and the two sides can communicate at any time, and the real-time is strong. 3 The data is transmitted in sequence in the transmission order during communication, and there is no sequerice issue. 4 Circuit Exchange is applicable to transmitting analog signals and is also suitable for transmitting digital signals. 5 switching of switching equipment (switches, etc.) and control are simpler. Disadvantages: 1 The average connection settling time of circuit exchange is suspects to communicate computer communication. 2 After the circuit switched connection is established, the physical pathway is exclusive, even if the communication line is idle, the channel is not used, so the channel is low. 3 Circuit exchange, data direct, different types, different specifications, different speeds of different rates, difficult to communicate with each other, and it is difficult to perform errors in communication. (2) Packet exchange: packet exchange is a unit of packets, packets carrying target addresses, source address, etc., using storage forwarding mode, so there is the following advantages and disadvantages: Advantages : 1 Packet exchange does not require a dedicated communication line for communication between communication, and there is no connection to establish a connection, and users can send messages at any time. 2 Due to the transmission of storage forwarding, it has the following advantages: a. Easy to set up code inspection and data retransmission facilities in the message exchange, plus the exchange node also has path selection, you can do a transmission path. When the fault is faulty, another path transfer data is reselected, enhance the reliability of the transmission; b. Easy to implement code conversion and rate matching in storage forwarding, and even the two sides of the transceiver can be available at different states. This makes communication between the type, specifications, and speeds of different computers; c. Provide multi-target services, that is, a message can be sent to multiple destination addresses simultaneously, which is difficult to implement in circuit exchange; D. Allow Establish a priority of data transfer to give priority high-priority packets. 3 Communication Both are not fixed to a communication line, but a part of a part of a different period of time occupies this physical pathway, which greatly improves the utilization of communication lines. Disadvantages: 1 Since the data is subjected to the storage, the process is subject to the process, the process is forwarded, causing forwarding delays (including receiving packets, checking correctivity, queuing, transmission time, etc.), and the greater communication volume, resulting The greater time delay, so the real-time nature of the exchange of communication is not suitable for transmitting real-time or interactive services. 2 The message exchange is only available for digital signals. 3 Since the packet is not limited, each intermediate node must be completely received throughout the message. When the output line is not idle, it may also be stored several full packets waiting forwarded, requiring each of the network. There is a large buffer with a larger buffer. In order to reduce costs, reduce the capacity of the cushioning memory, sometimes on the packets waiting forwarded, further increase the transmission delay. (3) Packet switching: Group exchange still uses storage forwarding transmission mode, but divides a long message to a few shorter packets, then send these packets (carrying sources, destination addresses, and number information) one by one. Therefore, in addition to the advantages of packets, packet exchange has the following advantages and disadvantages compared to the packet exchange: Addition: 1 Accelerate the transmission of data in the network.

Since the packet is transmitted one by one, the storage operation of the latter packet can be parallel to the forwarding operation of the previous packet, which reduces the transmission time of the message. Further, the buffer required to transmit a packet is much smaller than the buffer required to transmit a message, so that the chance to send by the buffer is not enough, and the waiting time must be much less. 2 Simplify storage management. Because the length of the packet is fixed, the size of the corresponding buffer is also fixed, and the management of memory in the exchange node is often simplified to be managed to the buffer, relatively easier. 3 Reduce the error rate and the amount of data returned. Because the packet is short, its error rate is inevitably reduced, and the amount of data resended is greatly reduced, which not only improves reliability, but also reduces transmission delay. 4 Since the packet is short, it is more suitable for adopting a priority policy, which is convenient for timely delivery of emergency data, so a burst data communication between the computer is obviously more appropriate. Disadvantages: 1 Although the packet exchange is less than the transmission of the message exchange, there is still a storage forwarding time, and its node switches must have stronger processing capabilities. 2 Packet exchange is the same as the packet exchange, each packet is added to information such as the source, destination address, and packet number, so that the amount of information transmitted by about 5% to 10% is reduced to a certain extent, and the communication efficiency is reduced. Treatment time, make control complex, and increased delays. 3 When the packet exchanges uses a datagram service, there may be sequential, lost or repeated packets. When the packet arrives at the point of origin, the packet is sorted by numbers, and add trouble. If a virtual circuit service is used, although there is no sequencing problem, there is a call setup, data transfer and virtual circuit release three processes. In summary, the amount of data to be transmitted is large, and its transmission time is much larger than the call time, the circuit exchange is preferred; when the end-to-end path has a lot of segments, it is preferred to use packet switched transmission data. From the improved channel utilization of the entire network, the packet exchange and packet exchange is superior to circuit exchange, where the packet exchange is small than the time delay of the packet, especially for burst data communication between the computer.

5-05 Try to compare the advantages and disadvantages of the two services of virtual circuits and datagrams. Answer: (1) In the transmission mode, the virtual circuit service is in the source, the destination host communication, should first create a virtual circuit, and then communicate, and the communication should be removed. And the data newspaper service, the network layer receives packets from the transportation layer, put it on the header (source, destination address, etc.), as a separate information unit, does not need to establish and release the connection, the target node receives the data There is no need to send confirmation, and thus is a small development method. But the hairper does not know if the other party is ready to receive, whether it is busy, so the reliability of the data newspaper service is not very high. (2) About the full network address: The virtual circuit service only needs to fill in the entire network address of the source and destination host only in the source host, and only fill in the virtual circuit number in the data transfer phase. And the data newspaper service, since each datagon is transmitted separately, there must be a full network address of the source and destination host in each datagram, so that the network node is forwarded according to the host with the address, which is frequent. People - Machic Interaction Communication At each time, the whole network address of the destination host is not only cumbersome, but also reduces the channel utilization. (3) Regarding routing: The node along the virtual circuit service is only transmitted when the call request packet is transmitted in the network, and it will not need it in the future. However, in the data newspaper service, each data must be routed once every network node is passed. When there is a long message that needs to be transmitted, it must first divide it into a number of packets with fixed length. If the datagram service is used, it will increase the network overhead. (4) About the packet sequence: For the virtual circuit service, since all packets from the source host are transmitted by the prior establishment of a virtual circuit, the packet can be guaranteed to reach the host in the sequence of sequence. However, when a long message is divided into several short datasuna, since they are independently transmitted, they may each pass different paths to the host, so the data newspaper service cannot guarantee that these datagrams reach the host. (5) Reliability and adaptability: The virtual circuit service has been connected before communicating, and each other after a certain number of packets, the other party also gives a confirmation, so the virtual circuit service is higher than the reliability of the data newspaper service. However, when a node or link in the transmission is faulty, the data newspaper service can bypass these faults, while other paths, transfer the data to the destination, and the virtual circuit service must re-establish the virtual circuit. Can communicate. Therefore, the adaptability of the data newspaper service is better than the virtual circuit service. (6) About Balance Network Flow: Data News In the transmission process, the relay node can select a small route to the data report, and avoid the high traffic, so the data newspaper service is both balancing the information in the network. Traffic, but also makes the data report more quickly. In the virtual circuit service, once the virtual circuit is established, the relay node is not possible to change the transmission path of the packet according to the flow conditions. In summary, the virtual circuit service is suitable for interaction, not only timely, transmission is relatively reliable, and the network overhead is small. Datashers 5-06 feature a communication subnet. If you use a virtual circuit, each packet must have a 3-byte grouping header, and each network node must reserve 8 bytes of storage space to identify the virtual circuit. However, if the datag is used, each group has a 15-byte group header, and the node does not need to retain the storage space of the routing table.

To set up to 0.01 yuan per pass, the fee is 0.01 yuan per word, and the life of the memory is 2 years working hours (40 hours a week). It is assumed that each average usage time of a virtual circuit is 1000 seconds, and 200 packets are transmitted during this time, and each packet is to pass through 4 links. Trial: Which program (virtual circuit or datagram) is used more economical? How much is the difference? A: 4 links mean to involve 5 routers. The virtual circuit implementation requires a memory that is fixedly assigned a 5 × 8 = 40 byte in 1000 seconds. Data reporting requires the capacity of multi-transmitted head information that requires multiple transmission than virtual circuits, equal to (15-3) × 4 × 200 = 9600 bytes, link. The current problem is a 40,000-byte · second memory comparison of 9600 bytes of circuit capacity. If the memory usage is two years, 3600 × 40 × 52 × 2≈1.5 × 107 bytes · seconds is 1 ÷ (1.5 × 107) = 6.7 × 10-8 points, then 40000 bytes · second The cost is about equal to 2.7 ms. On the other hand, 1 byte · link cost is 10-6 points, 9600 bytes · link cost is 10-6 × 9600 = 9.6 × 10-3 points, which is 9.6 ml. Obviously, for such parameters, the realization of the virtual circuit should be cheaper. 9.6-2.7 = 6.9 ml, that is, it is approximately 6.9 ml of this 1000 seconds. Suitable for transmitting a single packet, there is no interaction, and an occasion of not high transmission requirements.

5-06 has a communication subnet. If you use a virtual circuit, each packet must have a 3-byte grouping header, and each network node must reserve 8 bytes of storage space to identify the virtual circuit. However, if the datag is used, each group has a 15-byte group header, and the node does not need to retain the storage space of the routing table. To set up to 0.01 yuan per pass, the fee is 0.01 yuan per word, and the life of the memory is 2 years working hours (40 hours a week). It is assumed that each average usage time of a virtual circuit is 1000 seconds, and 200 packets are transmitted during this time, and each packet is to pass through 4 links. Trial: Which program (virtual circuit or datagram) is used more economical? How much is the difference? A: 4 links mean to involve 5 routers. The virtual circuit implementation requires a memory that is fixedly assigned a 5 × 8 = 40 byte in 1000 seconds. Data reporting requires the capacity of multi-transmitted head information that requires multiple transmission than virtual circuits, equal to (15-3) × 4 × 200 = 9600 bytes, link. The current problem is a 40,000-byte · second memory comparison of 9600 bytes of circuit capacity. If the memory usage is two years, 3600 × 40 × 52 × 2≈1.5 × 107 bytes · seconds is 1 ÷ (1.5 × 107) = 6.7 × 10-8 points, then 40000 bytes · second The cost is about equal to 2.7 ms. On the other hand, 1 byte · link cost is 10-6 points, 9600 bytes · link cost is 10-6 × 9600 = 9.6 × 10-3 points, which is 9.6 ml. Obviously, for such parameters, the realization of the virtual circuit should be cheaper. 9.6-2.7 = 6.9 ml, that is, it is approximately 6.9 ml of this 1000 seconds.

5-07 Assume that all nodes in the communication subnet work normally, all software is correct. Question a group that may be delivered to the wrong point of the error (no matter how small this probability)? If the data link layer protocol in a network can work correctly, is it necessary to be reliable from the end-to-end communication between the source node to the destination node? A: (1) It is possible. Large burst noise may destroy the group. Using K bit tests and error, there is still 2-k probability of leaks. If the packet's destination address segment or the virtual circuit number is changed, the packet will be delivered to the wrong destination and may be received as the correct packet. In other words, accidental burst noise may change the full legitimate packet to a destination into another destination is also a completely legally packet. (2) End-to-end communications is not reliable. End-to-end communication is not only related to the data link layer, but also related to the network layer, although the link layer protocol works correctly, but does not guarantee the normal work of the network layer protocol, that is, whether the communication subnet is reliable. What is the main agreement of 5-12 X.25? What is its characteristic? A: X.25 is a specifications for the public packet switched network interface. The interior of the network is determined by each network. "X.25 Network" only describes the interfaces of the network and network external data terminal devices to follow the X.25 standard. X.25 is based on virtual circuit service. The X.25 interface is divided into 3 levels. The lowermost physical layer interface standard uses X.21 Recommendation, the most used is RS-232 standard. The second layer is the interface standard used by the data link layer, which is a balanced link to access LAPB, which is 1 subset of HDLC. The third layer is a grouping layer, and multiple logical channels (0 --- 4095) can be created between DTE and DTC, and 1 DTE can establish multiple virtual circuits simultaneously to communicate. X.25 Group exchange data network features multi-communication, greatly improve line utilization, information transmission security, reliable, high transmission rate, through application account, identification (NUI), can achieve national roaming, provide rate from 2400b / s To 64kb / s.

6-02 The first inspection in the IP datagram and does not check the data in the datagram, what is the biggest benefit of doing this? What is the bad place? A: The error in the head is more serious than the error in the data. For example, a bad address may result in packets being sent to an error. Many hosts do not check if the group delivered to them is really to be delivered. They assume that the network will never deliver the group to them to another host. Sometimes data does not participate in the inspection and calculations, because this cost is large, and the upper protocol usually does this test work, causing repetition and extra. Therefore, this can speed up the forwarding of the packet, but when the data portion occurs, it cannot be discovered early.

6-03 A 3200bit long TCP packet is transmitted to the IP layer, and the first part of 160bit has become a datam. The following internet is connected by a router by two local area networks. However, the data portion in the longest data frame that the second local area network can only have 1200 bits. Therefore, the datagram must perform shards in the router. Speaking about how many bits of bits in the second LAN to its upper layer ("data" herein, of course, is the data seen by the local area network)? A: Entering the IP layer when the IP layer is 3200 160 = 3360bit; after the network layer of the local area network, add two head information, at this time, there is a total of 3360 160 160 = 3680bit; in the second local area network, The packet is to be slice, and the data portion of the longest data frame is only 1200bit, so it is divided into 4 pieces, so the second LAN will transmit 3840 bit.

6-04 Some units are assigned to a Class B IP address with NET-ID 129.250.0.0. This unit has more than 4,000 machines distributed in 16 different locations. If you select subnet mask 255.255.255.0, give each site allocated a subnet number and calculate the minimum or maximum of each location host number. A: The minimum value of the host number of each location is 1, the maximum value of 254.6-05 a hexadecimal representation of an IP address is C22F1481, and the trial will be converted into a dichotomy ten. Which IP address is this address? A: According to the decimal decade, the IP address is 194.47.20.129, which is Class C address.

6-06 Some people think: "The ARP protocol provides the service service to the network layer, so the ARP should belong to the data link layer." Why is this wrong? " A: ARP does not provide services to the network layer, which itself is part of the network layer and helps to provide services to the transport layer. There is no IP address in the data link layer. The data link layer protocol is a protocol like HDLC and PPP, which transmits a bit string from one end of the line to the other end.

6-07 ARP and RARP are mapped to another space from one space to another. In this sense, they are similar. However, ARP and RARP are very different in terms of implementation. Please point out this difference. A: A RARP server is responsible for answering the query request in the implementation of RARP. There is no such server in the implementation of the ARP, and the host answers the ARP query yourself.

6-08 Subnet mask of a Class B address on the Internet is 255.255.240.0. How much is the number of hosts on each subnet? A: For a Class B network, the high-end 16-bit forming the network number, the low end 16 is the subnet or host domain. In the low end 16 bits of the subnet mask, the maximum effective 4 bits are 1111, so the remaining 12 bits (4 bits and 4th bytes) are left for host numbers. Therefore, there are 4096 host addresses, but because all 0 and all 1 are special addresses, the maximum number of hosts should be 4094.

6-09 There is a "protocol" field in the IPv4 first, but there is no in the fixed first part of IPv6. Why is that? A: The purpose of setting the protocol field is to tell the destination owner Which protocol handler is handed over to the IP packet. The router in the middle does not need this information, so you don't have to put it in the main head. In fact, this information exists in the main head, but being disguised. The next header of the last (extension) head is used for this purpose.

6-10 When using IPv6, whether the ARP protocol needs to change? If you need to change, you should change conceptual changes or technical changes? A: From the concept, it is not necessary to change. In terms of technology, since the requested IP address is getting bigger, there is a large domain (also called section).

6-11 IPv6 uses 16-byte address space. Set up 1 million addresses every 1 slightly. How many years should I use to calculate how many years can be used up in the IP address space. It can be compared with the age of the universe (approximately 10 billion years). A: Using 16 bytes, the total number of addresses is 2128 or 3.4 × 1038. If we allocate them at a rate of 1018 per second, 1018, these addresses will last 3.4 × 1020S, which is about 1013. This figure is 1000 times the universe age. Of course, the address space is not flat, so their allocation is non-linear, but this calculation results indicate that even if the allocation scheme is one-thousandth, so many addresses will never be used.