In IEEE 802.11 infrastructure networks, access points (APs) send beacons periodically to stations (STAs), about once per 0.1s which is configurable. Beacons serve for time synchronization and power management. In multi-hop ad networks, beacons may be lost, incur too much overhead or cannot achieve their desired functions. In this project, you need to implement the beaconing scheme in IEEE 802.11 distributed coordination function (DCF), evaluate its influence on the normal operations of data and routing traffic in multi-hop networks and propose better schemes. (Currently, both GloMoSim and ns2 do not simulate the beaconing scheme.)

Two papers (MMAC by So and Vaidya and SSCH by Bahl et. al) propose MAC protocol that uses multiple channels with single transceiver. In this project, you are required to implement either these two multi-channel MAC protocols or one of them and and compare their performance with the single-channle MAC already implemented in either ns-2 or GLoMoSim.
Because MMAC and SSCH assume time synchronization and it will be good if you can come up with proposals that do not require time synchronization.

IEEE 802.11e working group has proposed QoS enhancements for IEEE 802.11 based wireless LANs and the draft is available from the class web site. In 802.11e, stations maintain separate queues for different classes of traffic. Without queue management, some queues may be starved especially for those queues holding asynchronous traffic. It is also bad when certain flows monopolize the channel and congestion happens. Besides, in multi-hop networks, priority reversal may happen (a low-priority packet from one station may preempt a high-priority packet from another station) due to hidden terminals or location dependent contention. In this project, you need to propose some congestion control and queue management schemes for IEEE 802.11e based networks. Hint: IEEE 802.11e simulation code is available from ns2 2.26 which is from TKN, a university in Germany.

The goal of the project is to build a physical layer emulation module in a standard linux IP stack. The use of such a module would be to emulate physical layer characteristics of mobile ad hoc networks. The module has to export a standard physical layer interface such as carrier sense, and transmit/receive packets to the wireless MAC driver (say, atheros 802.11 drivers for linux). The module will require an input/output interface with the MAC layer and the IP layer. Subsequently, the project can be extended to encapsulate standard IP packets going through the wireless interface and redirecting them through a wired interconnection network depending on a pre-defined topology. The glomosim radio layer can be used as a base for the physical layer emulation module. Additional information needs to be passed in the encapsulated packets to perform the actual emulation.

This project will involve extending either DSR or AODV to do multipath routing. DSR and AODV are on-demand protocols in which routes are established during a "Route Discovery" phase. The route discovery process involves flooding "Route Requests" and receiving "Route Replies". In the standard versionsof these protocols only the "shortest" route is chosen from the different Route Replies received (based on the shortest route or the route with the latest sequence number).

In the multipath version, you will modify the routing protocol to maintain all routes to destinations (instead of just one). You will implement two versions of multipath routing: one that only discovers totally (link) disjoint paths and another that relaxes that constraint and finds all paths (disjoint or not).
All routes will have to be used for subsequent data transmissions. To decide which route to use for transmitting data, you can use a simple probabilistic method or plain round-robin. In the probabilistic method you can assign a probability for using each route based on route metrics such as number of hops, delay etc. In the round-robin method the first packet will traverse the first route and the second packet will traverse the second route and so on.
For the project, students will have to compare the performance of the multipath versions of DSR or AODV with the standard version and turn in a report detailing the performance differences in terms of routing overhead and packet delivery ratios.

For this project, students will compare the performance of TCP-F (feedback) with regular TCP over wireless links. In wireless networks, most packet losses occur due to bit errors and routing failures on account of mobility. However regular TCP considers these losses as indicators of congestion and enters the congestion control phase and reverts to slow-start phase affecting performance. In the modified version called TCP-F the source is informed of route failures so that it does not invoke congestion control and refrains from sending more packets till the route is restored. (Pls read the paper : A Feedback Based Scheme For Improving TCP Performance In Ad-Hoc Wireless Networks, for details about the implementation).

In terms of the project, students will have to modify TCP to TCP-F with feedback mechanism. You can use either DSR or AODV as the routing protocol and compare the performance of TCP-F with TCP for different applications (ftp, http) under different mobility scenarios.

Use a sensor network as a communication channel. There is a sensor network deployed, there are "users" of the sensor network moving around the area. They want to communicate with each other (like walkie-talkies). Design and test a protocol to allow for such communication where there might be more than 2 parties involved.