QoS-Based Rate Control Scheme for Non-Elastic Traffics in Distributed Networks

Abstract

Non-elastic traffics are typically sensitive to assigned bandwidth and transmission delays which need quality-of-service (QoS) support to satisfy their different bandwidth requirements. QoS-based bandwidth allocation can be easily realized in a centralized manner for networks with central controllers. Since the centralized manner needs a central node to assign bandwidth, it is not suitable for distributed networks that do not have central control points. Thus, a distributed end-to-end bandwidth allocation mechanism is needed for distributed networks. We propose an end-to-end rate control scheme, called QoS-Based Rate Control (QBRC), for non-elastic flows in distributed networks. QBRC is implemented in transport layer and uses additive-increase multiplicative-decrease (AIMD) congestion control approach. QBRC has a capability to allocate bandwidth in a distributed manner. By adjusting AIMD parameters of the flows based on their bandwidth and jitter requirements, QBRC can implement the bandwidth allocation in the sender side and no central node is needed. Simulations show that QBRC is of great potential for differential non-elastic traffics in distributed networks.

Overload control in the network domain of LTE/LTE-A based machine type communications

“…Here, the eNB action is based on Additive Increase Multiplicative Decrease (AIMD) rule when the congestion or resource underutilization condition in the CN is detected. AIMD rule is one of the most popular approaches for controlling the sending rates through the binary feedback control message [21,23]. The AIMD based approach for adjusting N 2 L;i is given by (16).…”