QoS aware transmission control scheme for multihop WiMAX network

Abstract

The IEEE 802.16j Mobile Multihop Relay (MMR) WiMAX network allows the number of hops between the end user and the base station to be more than two hops. It supports non-real-time Polling Service, which considers the minimum reserved rate and the maximum sustained rate as a QoS requirements. The reliability of sending the data over MMR WiMAX is achieved by using Transmission Control Protocol (TCP) in transport layer and automatic repeat request in the link layer. However, the use of automatic repeat request in the link layer makes the round trip time fluctuate rapidly, which increases the possibility of retransmission timeout (RTO) expiration. TCP performance degrades because of frequent timeout, and hence the QoS transmission rates cannot be satisfied. Therefore, this paper presents an RTO smoothing scheme and QoS aware transmission control to enhance the performance of data transmission over MMR WiMAX networks. The RTO smoothing scheme aims to reduce the frequent timeout occurrences. The slow start threshold and maximum congestion window are adjusted to satisfy the required QoS and it provides transmission rate fairness for the users at different hops. The results showed that, the proposed schemes reduce the timeout, and improve the utilization of the allocated resources and TCP throughput.

A queuing model for distributed scheduling in IEEE 802.16 wireless mesh networks

“…Frequently, cross-layer approaches are proposed for mesh networks, as in the cross-layer approach where the information of the physical and link layers are shared with the upper layers in order to allocate network resources with relevant and updated information. Ismael et al [16] integrate information from the transport layer with the link layer, calculating the retransmission time in the link layer at an earlier time to reduce timeouts and enhance QoS transmission rates. Ismael et al [16] integrate information from the transport layer with the link layer, calculating the retransmission time in the link layer at an earlier time to reduce timeouts and enhance QoS transmission rates.…”

“…Ding et al [12] and Shakkotai et al [15] propose the cross-layer approach to integrate channel assignment tasks from the link layer with routing tasks from the network layer. Ismael et al [16] integrate information from the transport layer with the link layer, calculating the retransmission time in the link layer at an earlier time to reduce timeouts and enhance QoS transmission rates. Ye et al [17] present another treatment, integrating the link layer and the transport layer; the link layer gets explicit congestion notification messages from upper layers and makes decisions such as prioritizing some flows to reduce congestion.…”

“…16 and IEEE 802.11 s-a mesh protocol from the WiFi family. A lot of studies address the Point-to-Multipoint (PMP) case-where the stations talk directly with the MBS.…”

“…When designing an IEEE 802. 16 network, we must answer at least one of these following questions: (i) What is the guaranteed throughput? (ii) What is the delay for a packet to reach the MBS coming from a node n hops away from the MBS?…”

“…Studies using queuing theory for mesh networks generally do not incorporate the specific MAC of IEEE 802.16 standard and, thus, do not allow comparative analysis of standards, such as between IEEE 802. 16 and IEEE 802.11 s-a mesh protocol from the WiFi family. Therefore, the main contribution of this article is an analytical model based on queuing theory, which represents the MAC of IEEE 802.16 in mesh mode with distributed scheduling.…”