Communication Networks (10)

Reliable Data Transfer: Intuition

Selective Repeat

• receiver individually acknowledges all correctly received pkts
• sender only resends pkts for which ACK not received
• sender window

Selective repeat: dilemma

Receiver can’t see sender side. Assume using Seq. No. 0, 1, 2, 3.

TCP

point-to-point

one sender, one receiver

reliable, in-order byte steam

no “message boundaries”

pipelined

TCP congestion and flow control set window size

full duplex data

bi-directional data flow in same connection

MSS (maximum segment size)

connection-oriented

handshaking (exchange of control msgs) inits sender, receiver state before data exchange

flow controlled:

sender will not overwhelm receiver

Sequence number & ACK number

sequence numbers

byte stream “number” of first byte in segment’s data

acknowledgements

seq # of next byte expected from the other side

cumulative ACK

Sequence numbers acknowledgements are 32-bits unsigned integers.

1. A sends ‘C’, which is the 42-th byte sent.
2. A has recieved the 78-th byte from B and expecting the 79-th.
3. B sends ‘C’, which is the 79-th byte sent
4. B has recieved the 42-th byte from B and expecting the 43-th.

Tcp round Trip Time, Timeout

RTT

SampleRTT

measured time from segment transmission until ACK receipt

EstimatedRTT

$$\text { EstimatedRTT }=(1-\alpha) \times \text { EstimatedRTT }+\alpha \times \text { SampleRTT }$$

typical value: $\alpha = 0.125$

Timeout

Timeout Interval

EstimatedRTT plus “safety margin”

larger variation $\to$ larger safety margin

Deviation in RTT:

$$\text { DevRTT }=(1-\beta) \times \text { DevRTT }+\beta \times | \text { SampleRTT }=\text { EstimatedRTT } |$$

Typeical $\beta = 0.25$

$$\text { TimeoutInterval }=\text { EstimatedRTT }+4 \times \text { DevRTT }$$