Date of Award
Doctor of Philosophy (PhD)
Harlan B Russell
Daniel L Noneaker
In mobile ad hoc networks transmission-scheduling channel-access protocols are of interest because they can ensure collision free transmissions and provide fair access to the channel. The time taken to gain access to the channel is deterministic and hence these types of protocols can also guarantee a certain quality of service. However, these protocols suffer from two major drawbacks. The first issue is poor utilization of the channel due to fixed slot assignments. Once the slot assignments are decided they are held constant for a period of time. As a result the node to which a slot is assigned may not always have a packet to transmit in its assigned slot. This
results in wasted slots and leads to poor utilization of the channel. The second issue is that there is no support for networks with rate adaptive radios. In this work a combined solution to both of these shortcomings is presented. In order to make transmission-scheduling channel-access protocols support networks with rate adaptive radios, a process called slot-packing is developed. The design
of slot-packing ensures that it works with any transmission-scheduling channel-access protocol. Using slot-packing, we design and investigate a new protocol called adaptive recovering mini-slot transmission scheduling (RMTS-a) that tackles both the shortcomings
and improves the performance of the network significantly. A key feature of our RMTS-a protocol is that if a radio assigned to a transmission opportunity is unable to utilize all of the time slot, other radios in the local neighborhood are given the opportunity to transmit in the remaining time. Additionally, because multiple radios within communication range of a transmitter are likely to be able to decode the payload, packets to multiple neighbors can be packed within a single transmission.
Bollapragada Subrahmanya, Vikas, "A Channel-Access Framework for Scheduling Transmission Assignments in Ad Hoc Networks with Rate Adaptive Radios" (2020). All Dissertations. 2680.