Abstract not found

Ad hoc networks formed without the aid of any established infrastructure are typically multi-hop networks. Location dependent contention and "hidden terminal" problem make priority scheduling in multi-hop networks significantly different from that in wireless LANs. Most of the prior work related to priority scheduling addresses issues in wireless LANs. In this paper, priority scheduling in multi-hop networks is discussed. We propose a scheme using two narrowband busy tone signals to ensure medium access for high priority source stations. The simulation results demonstrate the e#ectiveness of the proposed scheme.

IEEE 802.11 contains a mechanism for transmission of data with realtime constraints known as Point Coordination Function. This supplementary medium access protocol resides on top of the basic medium access mechanism Distributed Coordination Function and uses a centralized polling approach. Due to the complexity of a PCF implementation and the predicted ineciency of the PCF several proposals have been presented for providing QoS support without the need of a centralized scheduler. Those solutions suer from the fact that they are shifting implementation complexity from the access point to the mobile nodes. In this paper we compare the suitability of the basic DCF and PCF protocols for the transmission of audio data in an interactive scenario. We show that a simple priority mechanism used on the mobiles as well as the access point is suitable for providing improved QoS in terms of bandwidth and without the need of an extended DCF protocol. In combination with the PCF an adequate delay characteristic for audio ows is achievable as well. To overcome the limitations in channel capacity caused by the PCF we suggest an implicit signaling scheme for improving the channel capacity by avoiding unsuccessful PCF polling attempts. Keywords IEEE 802.11, WLAN, real{time, best{eort, voice transmission, scheduling, PCF, DCF 1.

This report proposes an adaptive service differentiation scheme for QoS enhancement in IEEE 802.11 wireless ad hoc networks. Our approach, called Adaptive Enhanced Distributed Coordination Function (AEDCF), is derived from the new EDCF introduced in the upcoming IEEE 802.11e standard. One of the main problems of EDCF is the static reset of the Contention Window (CW) which decreases significantly the throughput performance and increases the collision rate specially at high load conditions. Our scheme aims to share the transmission channel efficiently while providing dierent priorities to access the wireless medium. Relative priorities are provisioned by adjusting the size of the Contention Window (CW) of each traffic class taking into account both applications requirements and network conditions. We have

In recent years, fair scheduling and quality of service (QoS) in wireless LANs have received significant attention from the networking research community. This paper presents a distributed algorithm for priority and fair scheduling in a wireless LAN. The proposed protocol is derived from HIPERLAN which is a wireless LAN standard defined by ETSI. The proposed protocol supports multiple priorities, as well as a mechanism (using weights) for controlling the way bandwidth is shared by ows within a given priority level.

Priority and Fair Scheduling in Wireless Local Area Networks. (August 2001) Anurag Dugar, B.E., Karnataka Regional Engineering College, India Co--Chairs of Advisory Committee: Dr. Nitin H. Vaidya Dr. A. L. Narasimha Reddy In recent years, fair scheduling and quality of service (QoS) in Wireless Local Area Networks have received significant attention from the networking research community. This thesis presents a distributed Medium Access Control (MAC) protocol for priority and fair scheduling in a Wireless Local Area Network. The proposed protocol supports multiple priorities, as well as a mechanism (using weights) for controlling the way bandwidth is shared by flows within a given priority level. iv To Nani, Mummy and Papa v ACKNOWLEDGMENTS I would like to thank my advisor and co-chair Dr. Vaidya for all his guidance, encouragement and valuable comments. His enthusiasm towards research and teaching has been a great source of inspiration. He has been a great mentor and has always been there to o#er me advice regarding academics and career. Without his help and support, my graduate school experience would not have been so fulfilling. I would like to thank my co-chair Dr. Reddy for all his guidance and time. He has always been very supportive and helpful. I would like to thank Dr. Choi for being part of my advisory committee. I would like to thank the members of the Mobile Computing and Networking group for the numerous discussions we had in di#erent areas of Wireless and Mobile ad hoc Networks. Being a part of this group has greatly enhanced my knowledge and has opened up my mind to various research possibilities in the area of Wireless and Mobile ad hoc Networks. I would like to thank NSF and Microsoft Research for supporting me financially to carry out resear...

Due to the increasing popularity of multimedia information over the web, streaming of continuous media such audio and video has received a tremendeous attention. Such trac typically tolerates some loss but has rigid delay constraints. We want to investigate and develop trac management schemes for this kind of services over a shared channel, such as Ethernet. Our approach is to smooth each of the streams and statistically multiplex them onto the shared channel so that the loss are below a miniscule threshold while maintaining the delay constraint. The parameters of the smoother are obtained from the uid model analysis assuming the media stream as uid owing into a medium with a restrict capacity. For the scheduler we use strategies such as pure ALOHA, CSMA without retransmission and the standard Ethernet medium access control, CSMA/CD. We also investigate the tradeo between delay spent in the smoother and in the scheduler. From this we can conclude when it is bene cial to invest some delay in the smoother for trac management purpose.