Presentation is loading. Please wait.

Presentation is loading. Please wait.

2. Getting connected (part 2)

Similar presentations


Presentation on theme: "2. Getting connected (part 2)"— Presentation transcript:

1 2. Getting connected (part 2)
Rocky K. C. Chang Department of Computing The Hong Kong Polytechnic University 13 February 2017

2 1. Automatic repeat request (ARQ)
Approach to achieving link reliability: Ask for retransmission when a corrupted frame is detected. Mechanisms for asking a sender to retransmit: Receiver to send negative acknowledgments for corrupted frames, or Receiver to send positive acknowledgments for good frames. Which one is better?

3 1. Automatic repeat request (ARQ)
Both approaches require a timeout mechanism. Negative-ack approach: A timer starts when a negative acknowledgment is sent out. Positive-ack approach: A timer starts when a message is sent out. Retransmissions take place when timeouts occur: Negative-ack approach: The receiver retransmits a negative acknowledgment. Positive-ack approach: The sender retransmits a frame. When using the –ve ack, how does the receiver know about a packet loss, especially when the first packet is lost?

4 1. Automatic repeat request (ARQ)
The ARQ implements a positive acknowledgment approach. Only after an acknowledgment is received will a frame be removed from a send buffer. An acknowledgment can be piggybacked on a message sent to the other direction. An acknowledgment sometimes indicates the sequence number of the next expected frame. Positive acknowledgment is usually accumulative, e.g., receiving an acknowledgment for frame 4 implies that frames 1-3 are all received correctly.

5 1.1 Stop-and-wait ARQ The maximum number of unacknowledged frames is one. A sender cannot send a second frame before receiving an acknowledgment for the first frame. The minimum number of sequence numbers needed to identify the frames is two (0, 1), i.e., the first two in the send buffer. 0: for the frame sent and waiting for its acknowledgment. 1: for the frame to be sent after receiving an acknowledgment for the 0th frame.

6

7 1.1 Stop-and-wait ARQ Advantages: Disadvantages:
Very simple to implement, both on the sender and receiver sides. Disadvantages: Achieve a very low throughput, especially in a high-speed link (does not keep the pipe full). In the best case, only one frame can be sent in a round-trip time. In other cases, an additional number of round-trip times is required when errors occur.

8 1.2 Sliding-window (or go-back-n) ARQ
The maximum number of unacknowledged frames may be more than one. A sender may continue to transmit frames even when acknowledgments for frames previously sent are not received. A sliding window is used to keep track of the sender’s state and receiver’s state. The window size determines the maximum number of unacknowledged frames allowed on the sender side.

9 1.2 Sliding-window (or go-back-n) ARQ
On the sender side: Frames are labeled by a sequence number of integer values, starting from 1. A static parameter: send window size (SWS) Two variable parameters: The maximum sequence number being acknowledged by the receiver (LAR) The maximum sequence number sent by the sender (LFS) Note that LFS  LAR LFS = LAR: All frames sent have been acknowledged. LFS > LAR: LFSLAR frames are yet to be acknowledged.

10 1.2 Sliding-window (or go-back-n) ARQ
Also note that LFS  SWS + LAR. When LFS = SWS + LAR, the window is said to be full, and no more new frames can be sent before (LAR+1)th frame is acknowledged. How do the values of LFS and LAR change? LFS is initialized to 0, and it is incremented after each new frame is sent. LAR is initialized to 0, and it is shifted to the right according to the new sequence number(s) acknowledged by the receiver.

11 1.2 Sliding-window (or go-back-n) ARQ
Sender side: Receiver side: = SWS LAR LFS = RWS NFE

12 1.2 Sliding-window (or go-back-n) ARQ
On the receiver side: A static parameter: receive window size (RWS) RWS is the maximum number of out-of-order frames that the receiver is willing to accept. A variable parameter: Sequence number of the next frame expected (NFE) If a frame, whose sequence number is larger than NFE+RWS, is received, the frame will be discarded; otherwise, it will be buffered in a receive buffer. When a frame is received correctly, NFE is either updated or unchanged.

13 1.3 Three uses of ARQs Provide reliable service
Provide in-order service The original order of the frames is preserved even when out-of-order frames are received. Provide flow control Prevent the sender from flooding the receiver. The sender needs to vary the value of SWS by taking into account of the state of the receive buffer.

14 2. Local area networks (LANs)
Early 1980s IBM’s token ring vs. DIX (Digital, Intel, and Xerox) Ethernet IEEE (logical link control), (Ethernet), (Token Bus), (Token Ring) Late 1980s Fiber Distributed Data Interface (FDDI) Distributed Queue Dual Bus (DQDB) Early 1990s ATM LANs vs Fast Ethernet (switched Ethernet)

15 2. Local area networks (LANs)
Mid-1990s IEEE (wireless LAN) 2000 Gigabit Ethernet Metro Ethernet Optical Ethernet Wireless Ethernet Development trends: From shared medium to switched LANs From router-based backbone to switched backbone From wired to wireless From single-medium to multi-media

16 3.1 Ethernet LAN: physical connectivity
Components: Cable (passive) Transceivers (transmitter + receiver) Adaptor (active). Each adaptor card is uniquely identified by a 48-bit (physical or MAC) address, e.g., 00:40:26:5A:67:88. Design principles: Cost-effective resource sharing Reliability Inexpensive

17 3.1 Ethernet LAN: physical connectivity
ransceiver Ethernet cable Adaptor Host

18 3.1 Ethernet LAN: physical connectivity
Both DIX and IEEE Ethernets do not require switching elements. Hosts are connected to a cable (10base2/5/T) through network adaptors. Several segments may be connected (horizontally) to another segment (vertically) through hubs, which serve as repeaters.

19 3.1 Ethernet LAN: physical connectivity
Repeater Host

20 3.1 Ethernet LAN: physical connectivity
from the datalink layer and up

21 3.2 Ethernet LAN: Datalink sublayers
A new multiple access control (MAC) problem: How do multiple hosts share a single transmission medium efficiently? This problem occurs in token ring, FDDI, and wireless LAN. An additional MAC sublayer was created for this purpose. A logical link control (LLC) sublayer: Provide similar services as a datalink layer except that error detection is provided at the MAC sublayer.

22 3.2 Ethernet LAN: Datalink sublayers
Provide three types of services: Unacknowledged connectionless (datagram) service: Basically no additional service. Acknowledged connectionless service: Reliability through a stop-and-wait-ARQ-like mechanism. Connection-mode service: A connection is set up between two hosts with flow control and reliability services.

23 3.2 Ethernet LAN: Datalink sublayers
The datalink layer consists of LLC sublayer and MAC sublayer IEEE Logical Link Control (LLC) Sublayer ANSI FDDI IEEE 802.3 IEEE 802.4 IEEE 802.5 IEEE IEEE

24 1-byte start frame delimiter
3.2 Ethernet LAN: frames DIX Ethernet frame structure: The 7-byte preamble is sent before the frame to allow the receiver to synchronize with the signal. IEEE Ethernet frame structure: 1-byte start frame delimiter 7-byte preamble 6-byte dest address 6-byte src address 2-byte type 4-byte CRC Data type 0800 IP datagram 802.3 MAC 802.2 LLC 802.2 SNAP dest address src address Preamble len DSAP AA SSAP AA cntl 03 org code 00 type 4-byte CRC Data

25 Ethernet frame

26 3.3 Ethernet LAN: MAC protocol
Types of MAC addresses: Unicast address: hardwired into ROM Broadcast address: all 1 bits Multicast address: First bit set to 1 and configurable. Promiscuous mode CSMA/CD (carrier sense multiple access with collision detection) Each adaptor is able to distinguish a busy link from an idle link. Each adaptor is able to detect “frame collisions,” if occurred, as it transmits.

27 3.3 Ethernet LAN: MAC protocol
To send a frame, transmits it immediately when the link is detected idle. The maximum length of the payload is 1500 bytes (jumbo frames with 9000 bytes) To receive a frame, Every adaptor attached to the link will receive a copy of a frame transmitted on the link. The frame will be discarded if the destination address does not match its unicast address, broadcast address, and any configured multicast addresses.

28 3.3 Ethernet LAN: MAC protocol
Carrier sense is not sufficient for avoiding frame collisions. A frame must be long enough to detect collisions: A sufficient condition: A frame occupies the entire pipe back and forth. For example, for a 10-Mbps Ethernet segment of 2500m long (a round-trip propagation delay of 51.2 s): Minimum frame length = 51.2s  10Mbps = 512 bits (64 bytes), or a 14-byte header + a 46-byte payload + a 4-byte CRC.

29 3.3 Ethernet LAN: MAC protocol
B (a) A B (b) A B (c) A B (d)

30 3.3 Ethernet LAN: MAC protocol
When more than one adaptor transmits frames “almost at the same time,” The frames are “collided” and can be detected by the adaptors involved. The adaptors involved then send a 32-bit jamming sequence, and stop transmission. The adaptors use exponential backoff for retransmission (up to a limited number of attempts). After first collision: either 0 or 51.2 s. After second collision: 0, 51.2, 102.4, s. After nth collision: k51.2 s for k = 0..2n1.

31 3.4 Ethernet’s performance
Throughput decreases with a and the number of hosts, where a = propagation delay/transmission delay. Implications: Limit on the number of stations and the maximum length of the Ethernet segment Effect of increasing link’s data rate Support for delay-sensitive data Methods of improving its performance Each collision domain consists of only one host.

32 The University of Adelaide, School of Computer Science
3 July 2018 4. Wireless Links Wireless links transmit electromagnetic signals Radio, microwave, infrared Wireless links all share the same “wire” (so to speak) The challenge is to share it efficiently without unduly interfering with each other Most of this sharing is accomplished by dividing the “wire” along the dimensions of frequency and space Exclusive use of a particular frequency in a particular geographic area may be allocated to an individual entity such as a corporation. Chapter 2 — Instructions: Language of the Computer

33 4. Wireless Links A wireless network using a base station

34 The University of Adelaide, School of Computer Science
3 July 2018 4. Wireless Links Mesh or Ad-hoc network Nodes are peers Messages may be forwarded via a chain of peer nodes A wireless ad-hoc or mesh network Chapter 2 — Instructions: Language of the Computer

35 The University of Adelaide, School of Computer Science
3 July 2018 4.1 Wireless LAN (IEEE ) Also known as Wi-Fi Like its Ethernet and token ring siblings, is designed for use in a limited geographical area (homes, office buildings, campuses) Primary challenge is to mediate access to a shared communication medium – in this case, signals propagating through space supports additional features power management and security mechanisms Chapter 2 — Instructions: Language of the Computer

36 4.1 Wireless LAN (IEEE 802.11) MAC problem:
Hidden node (terminal) problem: Frame collision occurs but the senders involved are unaware of it. AB and CB (A and C are not connected) Exposed node problem: A node is unnecessarily prevented from transmitting frames. BA and CD (A and D are not connected) Both problems are due to the fact that, unlike Ethernet, the nodes are not always connected together.

37 4.1 Wireless LAN (IEEE )

38 4.1 Wireless LAN (IEEE ) Multiple Access with Collision Avoidance (MACA) To send a frame, the sender first sends a Request to Send (RTS) frame to the receiver. The receiver then replies with a Clear to Send (CTS) frame back to the sender. Any node that sees the CTS frame will refrain from sending any frames (for the hidden node problem). Any node that sees the RTS frame but not the CTS frame is free to transmit (for the exposed node problem). When two or more nodes transmit an RTS frame at the same time, their RTS frames collide. Retransmission of RTS frames takes place after the nodes involved do not receive the CTS frames, i.e. no direct collision detection supported.

39 Conclusions We examined the five major problems for a network link.
The four problems apply to all links. The last problem applies to only broadcast media, e.g., LAN. We explained how the CSMA/CD protocol work in Ethernet. There are maximum and minimum frame size. Due to the CSMA/CD protocol, Ethernet cannot scale to number of hosts, maximum length of Ethernet segment, and data rate. For wireless LANs, there are additional problems, e.g., Interference Hidden terminal and exposed node problems CSMA/CA protocol to resolve the problems Error-correction code (for satellite networks) Fast handover (for mobile networks)

40 Acknowledgments The content in this set of slides is based on Peterson and Davie.


Download ppt "2. Getting connected (part 2)"

Similar presentations


Ads by Google