一、选择题(共 30 题,每题 1 分)
1. 局域网(LAN):概念,划分。(11 页,297 页) Ethernet, wireless LAN 此题答案不明确
2. 域名解析服务(DNS)三层结构:根(root)DNS 服务器,顶级域(top-level domain)服务器,权威(authoritative) DNS 服务器。DNS 记录结构。离你最近的 DNS 服务器上存放着什么?(87 页,90 页)
root name server: contacts authoritative name server if name mapping not known, gets mapping, returns mapping to local name server
顶级域 (TLD) servers: responsible for com, org, net, edu, etc, and all top-level country domains uk, fr, ca, jp. 权威Authoritative DNS servers:主机名映射为IP地址的DNS记录 本地Local Name Server: caches mapping
3. TCP 连接的建立:三次握手(three-way handshaking),SYN 和 ACK 标志位,顺序号和确认号。(169 页)
4. SMTP: 推协议(push protocol),邮件服务器间投递邮件,不使用中间邮件服务器,25 端口,只能用 7 位 ASCII 码表示。(77 页)
SMTP: sending server to receiving server, port 25, messages must be in 7-bit ASCII, SMTP server uses CRLF.CRLF to determine end of message, SMTP: push
5. TCP/IP 协议族(TCP/IP protocol suite)
application: FTP, SMTP, HTTP transport: TCP, UDP network: IP, routing protocols link: PPP, Ethernet
层次 协议 应用层(application layer) 传输层(transport layer) HTTP SMTP FTP TCP UDP 网络层(network layer) 数据链路层(data link layer) IP RIP OSPF BGP ARP PPP 6. 流水线(pipelining):允许发送方发送多个分组而无需等待确认,但必须增加序号范围且缓存分组。(146 页) Pipelining: range of sequence numbers must be increased, buffering at sender and/or receiver 7. URL 的格式(format):<协议>://<主机名>:<端口>/<路径> (谢希仁第五版教材 238 页) 8. C/S 结构与 P2P 结构(7 页,53 页)
Client-server architecture: server: always-on host, permanent IP address, server farms for scaling(规模化的服务器场).
clients: communicate with server, may be intermittently(间歇) connected, may have dynamic IP addresses, do not communicate directly with each other(客户机相互之间不能直接通信).
Pure P2P architecture: no always-on server, arbitrary(任意的) end systems directly communicate, peers are
intermittently connected and change IP addresses, example: Gnutella, Highly scalable(可扩展性高) but difficult to manage 9. 分用(demultiplexing)和复用(multiplexing)(128 页)
Demultiplexing at rcv host: delivering received segments to correct socket
Multiplexing at send host: gathering data from multiple sockets, enveloping data with header (later used for demultiplexing)
10. 后退 N 帧协议(go-back-N):滑动窗口(sliding-window)协议。累积确认(cumulative acknowledgement):接
收方对目前收到的最后一个正确有序的分组发出确认,丢弃所有乱序分组。如果超时,发送方将重传所有已 发送但还未被确认的分组。接收窗口大小为 1。(147 页)
11. 端口号(port number)的范围:0-65535,周知端口号(well-known port number)0-1023,登记端口号(registered
port)1024-49151,短暂端口号(ephemeral port)49152-65535。(129 页,谢希仁第五版教材 183 页) 周知端口号:是保留给诸如HTTP和FTP之类的周知应用层协议的
12. TCP 顺序号(sequence number)和确认号(acknowledge number):顺序号是该报文段首字节的字节流编号,确
认号是期望收到的下一个报文段的顺序号。(157 页) 补:假定数据流由一个包含500 000字节的文件组成,MSS为1000字节,数据流的首字节编号0。TCP将为该数据流构建500个报文段。第一个报文段的序号被赋为0,第二个报文段的序号被赋为1000,以此类推。
从主机B到达的每个报文段中都有一个序号用于从B流向A的数据,主机A填充进报文段的确认号是主机A
期望从主机B收到的下一字节的序号。
13. 发送窗口:受制于流量控制和拥塞控制,拥塞控制中的拥塞窗口(congestion window)和阈值(threshold) (147 页,180 页,183 页)
sender limits transmission: LastByteSent-LastByteAcked < CongWin After 3 dup ACKs: CongWin is cut in half, window then grows linearly
after timeout event: CongWin instead set to 1 MSS, window then grows exponentially(指数), to a threshold, then grows linearly(线性)
14. TCP 报文段结构,各字段的名称、作用和值。TCP 报文段首部的长度范围:20-60 字节。(157 页) URG: urgent data
(generally not used)urg=1为紧急
数据
ACK:=1 确认号有效 PSH: push data now (generally not used) # bytes rcvr willing to accept
RST, SYN, FIN: connection estab (setup, teardown
commands)
Internet
RST=1 tcp需释放重建 checksum
Syn=1请求/接受报文 (as in UDP)
Fin=1 data发送完毕
15. 拥塞控制(congestion control):慢启动(slow start),拥塞避免(congestion avoidance),加性增乘性
减(AIMD)。(180-184 页) additive increase: increase CongWin by 1 MSS every RTT until loss detected multiplicative decrease: cut CongWin in half after loss
Slow Start: When connection begins, CongWin = 1 MSS, available bandwidth may be >> MSS/RTT, When connection begins, increase rate exponentially fast until first loss event congestion avoidance: After 3 dup ACKs: CongWin is cut in half, window then grows linearly. after timeout event:
CongWin instead set to 1 MSS, window then grows exponentially(指数), to a threshold, then grows linearly(线性) 在收到3个冗余的ACK后取消慢启动阶段的行为称为快速恢复(fast recovery) 16. 路由表的查询:最长前缀匹配。(207 页) Longest prefix matching(例题见207页)
17. 网络流量(network traffic)和吞吐量(throughput)。(28 页,174 页,184 页) Bottleneck link.
Average throughout: 0.75 W/RTT
18. RIP 协议:基于距离向量(DV)算法,以跳数(hop)作为代价度量(metrics)。(242 页,251 页) Routing Information Protocol: distance vector algorithm, distance metric: # of hops (max = 15 hops) RIP/UDP/IP
19. OSPF 协议:基于链路状态(LS)算法。(239 页,253 页)
Open Shortest Path First: uses Link State algorithm and Dijkstra’s algorithm, multiple same-cost paths allowed (only one path in RIP) OSPF/IP 20. BGP 协议:自治系统(AS)间的选路。(255 页) Border Gateway Protocol BGP/TCP/IP
21. 差错控制方法(error detection method):奇偶校验(parity check),校验和(checksum),循环冗余校验 (CRC),各自的特点和计算方法。(284 页)
Parity check: 偶校验:d个数据比特加1个比特中1的总数为偶数,奇校验:奇数个1。
在二维奇偶校验中可以检测到出现了单个比特差错的事实,而且可以利用列和行的小标来纠正。通常用于如音频CD这样的音频存储和回放设备。
Checksum:互联网检验和就是基于这种方法。所有的k比特进行求和,求和时有溢出就要被回卷,之后进行反码运算,得到检验和,在接收方将所有的k比特与检验和进行相加,如果无错就应该是全1。
CRC (polynomial code 多项式编码):计算机网络中广泛的应用。
22. 循环冗余校验的计算。(287 页)
23. 随机接入(random access)协议:时隙 ALOHA,ALOHA,CSMA,各自的特点。(292 页)
Slotted ALOHA: when node obtains fresh frame, transmits in next slot, if no collision(碰撞): node can send new frame
in next slot. if collision: node retransmits frame in each subsequent slot with prob. p(概率p) until success. 好:single active node can continuously transmit at full rate of channel, highly decentralized: only slots in nodes need to be in sync, simple. 不好:collisions, wasting slots, idle slots At best: channel used for useful transmissions 37% of time
Pure (unslotted) ALOHA: simpler, no synchronization, transmit immediately, collision probability increases even worse than slotted Aloha
CSMA (Carrier Sense Multiple Access): listen before transmit, If channel sensed idle: transmit entire frame, human
analogy: don’t interrupt others.
CSMA/CD (Collision Detection): easy in wired LANs, difficult in wireless LANs, human analogy: the polite
conversationalist.
24. ARP 的过程:广播(broadcast)发送 ARP 分组,单播(unicast)发送响应。(301 页)
A wants to send datagram to B, and B’s MAC address not in A’s ARP table, A broadcasts ARP query packet, containing
B's IP address, dest MAC address = FF-FF-FF-FF-FF-FF, all machines on LAN receive ARP query, B receives ARP packet, replies to A with its (B's) MAC address, frame sent to A’s MAC address (unicast), A caches (saves) IP-to-MAC address pair in its ARP table until information becomes old (times out), ARP is “plug-and-play” 25. IP 地址与 MAC 地址的映射(mapping):NAT,ARP,如何获得下一跳的地址?(228 页,300 页)
26. MAC 地址的格式,二进制到十六进制的转换。
MAC地址的长度为48位(6个字节),16进制表示,如08:00:20:0A:8C:6D(书上的表示法为08-00-20-0A-8C-6D)。其前6位08:00:20代表网络硬件制造商的编号,而后6位代表制造商所制造的某个网络产品(网卡)的系
列号。
此题考的是计算,应该是二进制到十六进制的转换。 27. 特殊的 IP 地址:
全网网络地址 0.0.0.0 广播地址 255.255.255.255
本机回送(loopback)地址 127.0.0.1 组播地址 224.0.0.1
私网地址 10.0.0.0---10.255.255.255 172.16.0.0---172.31.255.255 192.168.0.0---192.168.255.255 28. IP 数据报格式,各字段的名称、作用和值。哪些字段会随着数据报的转发而改变?(216 页)
total datagram
IP protocol version length (bytes) number
header length
(bytes) “type” of data max number remaining hops (decremented at
each router)
for
fragmentation/ reassembly
upper layer protocol to deliver payload to
版本号 四位(ipv4) 首部长度 4位(最大值60字节)区分服务 8位 总长度16位(首部+data之和的长度) 标识 16位(计数器,产生数据报标识) 标志(最低位MF=1表明后面有分片;中位DF=0允许分片 TTL 8位(可能通过路由器数量的最大值) 上层协议 8位(数据在上层使用什么协议6=TCP;17=UDP)
协议中值为6表明数据部分要交给TCP,值为17表示数据要交给UDP。
在每台路由器上都必须重新计算检验和并存放在原处,因为TTL字段及选项字段可能会改变。
29. TCP 平均吞吐量=0.75*W/RTT。(184 页)
30. 二进制指数退避算法(binary exponential backoff algorithm)(308 页)k={0,1,…,2^m -1}m=min(n,10)等待K*512比特时间(1bit=0.1us)
此题请看书。
二、填空题(共 10 题,每题 1 分)
31. 协议的定义:protocols define format, order of massages sent and received among network entities,
and actions taken on massage transmission, receipt.定义了在2个或多个通信实体之间交换的报文格式和次序,以及在报文传输接收或其他事件方面采取的动作。 32. 流量强度(traffic intensity)= La/R R=链路带宽 L=分组大小 a=平均分组到达速率(25 页)
La/R ~ 0: average queueing delay small