Random-access MIMO in single hop and wireless mesh network settings

Hunchangsith, Konglit (2012). Random-access MIMO in single hop and wireless mesh network settings PhD Thesis, School of Information Technology and Electrical Engineering, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
s41512177_phd_thefinalthesis.pdf Thesis full text application/pdf 2.77MB 15
Author Hunchangsith, Konglit
Thesis Title Random-access MIMO in single hop and wireless mesh network settings
School, Centre or Institute School of Information Technology and Electrical Engineering
Institution The University of Queensland
Publication date 2012-01
Thesis type PhD Thesis
Supervisor Marius Portmann
Aleksandar Rakic
Total pages 176
Total colour pages 28
Total black and white pages 148
Language eng
Subjects 100510 Wireless Communications
100503 Computer Communications Networks
Formatted abstract
Communication networks have become a critical factor for the world development. While many regions have enjoyed good communications services, there are still a great number of remote areas in almost every corner of the globe that have very limited access to such important resources. Wireless Mesh Networks (WMNs) have emerged as one of the most promising technologies for providing a practical, economical and robust backhaul network to reach those areas.

It is undeniable that the physical layer plays the most crucial role in defining the capacity and performance of communication systems. Multiple-Input-Multiple-Output (MIMO) technique is regarded as one of the most promising techniques for the physical layer that offers a significant advancement of performance without the additional demand for the already extensively-used radio spectrum. Such benefits are possible by combining MIMO with suitable access protocols. Distributed protocols (such as random-access protocols), are desirable for WMNs. MIMO that is capable of resolving mixing signals can be beneficial for random-access protocols that suffer greatly from signal collisions. Therefore, an important question to answer is how beneficial MIMO is when combined with random-access protocols in single-hop and WMN settings. This thesis will provide a suitable answer to this important question by creating analytical tools to quantify the potential of random-access and MIMO in single-hop and WMN settings. Moreover, this thesis identifies the potential problems of using MIMO with random-access and proposes novel techniques for alleviating them. To achieve these goals, the following works are carried out sequentially.

First, the probabilistic theory is applied to create single-hop MAC layer models for ALOHA and Slotted-ALOHA MIMO. These MAC layer models are used to determine the upper bound performances of the ALOHA and Slotted-ALOHA MIMO from the MAC layer perspective. It is found that the maximum capacities of the Slotted-ALOHA and ALOHA-MIMO protocols increase approximately linearly with increase in the number of receiving antennas.

Second, the MIMO channel models of the multi-user MIMO are combined with the MAC layer model, allowing the parameters from the physical and MAC layers to be evaluated simultaneously to determine the interdependency of the two layers. Unlike the one-to-one model, the multi-user MIMO channel model incorporates the interference between the users. The identical independent distributed (i.i.d) MIMO channel and also the Rician MIMO channel are considered. Also taken into consideration are the physical parameters, namely, channel correlation and antenna mutual coupling. Nodes completely surrounded by objects (non-line-of-sight (NLOS) transmission) and nodes with both NLOS and LOS are included. The results show that reducing the probability of transmission of the nodes with a high degree of LOS components can improve the overall system capacity. Furthermore, it is found that the greater the number of transmitting antennas used for the channel accessing, the higher the reduction in the probability of having successful transmissions and thus the lower the capacity. To overcome this deficiency, the Spatial Adjustment (SpA) strategy is proposed for Slotted ALOHA to better exploit the available degree of freedom and thus maximise the probability of having successful transmissions. When SpA is activated, each station uses one antenna for channel access so that the probability of having a successful transmission is maximised. If the channel access is successful, the receiver advises the transmitter(s) of what will be an appropriate number of transmitting antennas to be used in the adjustment mode. The obtained results show that SpA can significantly increase the system capacity.

Third, the analysis moves from the single-hop to the wireless-mesh network setting. To investigate WMNs, a novel analytical model of the wireless-mesh network is proposed. This WMN model is independent of the specific MAC or PHY layer technology. It captures the high-level characteristics of WMNs and can serve as a general tool for assessing each individual node throughput of WMNs. The model is validated using discrete event simulations in a commercially-available software, OPNET. The validation tests show that traffic saturation is largely responsible for the issue of fairness caused by the different node locations. In particular, the saturation of some nodes leads to a data packet loss that results in unfairness for the nodes further away from the gateway. This problem in WMNs is addressed in the next part of the thesis under the use of the random-access MIMO protocol. For this purpose, the complete analytical model of a WMN employing random-access MIMO is developed. The investigation using this model shows that increasing the number of receiver antennas can steadily increase the system’s capacity. However, under common max-min fairness, the outcome is a severe unfairness for the nodes further away from the gateway. The top-forwarder SpA technique is proposed for alleviating the location-dependant fairness issue. The technique allows the router that forwards traffic for the most number of users to adjust its number of transmitting antennas. The obtained results show that fairness can be improved.
Keyword MIMO
Wireless mesh network
Random access protocol
Cross layer
Additional Notes 32, 67-68, 72-75, 77-78, 80, 82-83, 99-101, 104, 106-108, 110, 130-131, 142, 144, 151

Citation counts: Google Scholar Search Google Scholar
Created: Fri, 29 Jun 2012, 09:22:20 EST by Mr Konglit Hunchangsith on behalf of Scholarly Communication and Digitisation Service