Question
1. Following two questions are related to congestion control in wired networks. Let us assume that size of buffer is infinite.
A. assumes "token bucket" mechanism for congestion control. Explain how to compute a queuing delay for example (from when a packet comes in a bucket to when it goes out). Note that, you do not have to use mathematical equations.
B. assumes "dynamic window management" for congestion control. Describe what problems arise under a large window size on connection created.
In IEEE 802.11, a node retransmits a packet if previous transmission of a packet is unsuccessful. If TCP over 802.11 is used, together protocols can generate a large number of retransmissions of a given packet to make sure its reliable transmission. Explain the reason that the retransmissions are need for both TCP and 802.11 even though these have same purpose.
2. The EDCF (superior DCF) in IEEE 802.11e maintains 4 different queues to support way in categories (AC) (See the sides). IEEE 802.11e explains only MAC and PHY layers. Still though upper layer as in IP layer can categorize characteristics of 4 different packets (Voice, Video, Best Effort, and Background), MAC layer does not check any category because of encapsulation. Describe how EDCF identifies 4 different AC of packets from an upper layer without reference of IP or TCP headers.
3. Following question is related to SPEED routing protocol-
Multiple speeds are computed in MAC layer. The forwarder with highest speed is selected on MAC layer feedback. This selection could not produce a shortest path. Clarify the reason that SPEED uses this approach in its place of other shortest or reliable path selections and advantage of this approach against shortest and reliable pathways