Efficient implementation of resource sharing strategies in a multi-user wireless
environment can improve the performance of a network significantly. In this thesis
we study various scheduling strategies for wireless networks and handle the problem
of opportunistically scheduling transmissions using channel aware schemes.
First we propose a scheme that can handle users with asymmetric channel conditions
and is opportunistic in the sense that it exploits the multi-user diversity of the
network. The scheme requires the users to have a priori knowledge of their channel
distributions. The associated overhead is limited meaning it offers reduced feedback
load, that does not scale with the increasing number of users. The main technique
used to shrink the feedback load is the contention based distributed implementation
of a splitting algorithm that does not require explicit feedback to the scheduler from
every user. The users find the best among themselves, in a distributed manner, while
requiring just a ternary broadcast feedback from the scheduler at the end of each
mini-slot. In addition, it can also handle fairness constraints in time and throughput
to various degrees.
Next we propose another opportunistic scheduler that offers most of the benefits
of the previously proposed scheme but is more practical because it can also handle heterogenous users whose channel distributions are unknown. This new scheme
actually reduces the complexity and is also more robust for changing traffic patterns.
Finally we extend both these schemes to the scenario where there are fixed thresholds,
this enables us to handle opportunistic scheduling in practical systems that can
only transmit over finite number of discrete rates with the additional benefit that full
feedback session, even from the selected user, is never required.
|Date of Award||Jul 2012|
|Original language||English (US)|
- Computer, Electrical and Mathematical Science and Engineering
|Supervisor||Mohamed-Slim Alouini (Supervisor)|
- Multi-user diversity