Abstract
In our paper, we propose a modification of queue based exponential scheduler that dynamically
improves the channel gain of far away users (by increasing their credits) at the cost of a small reduction
of the same for the nearby users(by reducing the credits), since users nearby anyway have a fairly high
gain. Results clearly suggest that the proposed method is able to maintain a constant average
throughput(always greater than that of the P.F scheduling) while it far outperforms the P.F and QueueBased
Exponential Rule Scheduler in the delay characteristics. The throughput fairness index was
interestingly observed to improve with increasing number of users while it worsens for the other two
schemes.
Users
Please
log in to take part in the discussion (add own reviews or comments).