You are here

Home › Research Interests

Wireless Networking

Research topics:

 

Dynamic channel allocation research  Prof Lajos Hanzo, Dr. Jonathan Blogh
 Jon's area of research to date has been the capacity study of GSM based cellular networks, using fixed and dynamic channel llocation schemes. Since the performance of a cellular network is generally limited by the levels of interference present, the effects of using adaptive antenna arrays at the basestations to reduce this interference were analysed.

Work has also been carried out to investigate the network performance benefits of power control and power control used in conjunction with adaptive modulation techniques. It was found that using just power control reduced the required transmission powers, hence extended battery life, whilst simulataneously enhancing the network performance. However, when used with adaptive modulation these benefits were compounded with improved modem throughput.

John's current work is focused on investigating the network capacity and performance of the next generation CDMA based, UMTS cellular networks due to come into service in a couple of years time.

Figure 1: Graph of minimum uplink SINR of the 7centre cells of a GSM based cellular network.

[top]
Ad Hoc wireless networks   Prof Lajos Hanzo, Kin C. Yong
What is an ad hoc network? An ad hoc network is a collection of mobile nodes without the need of infrastructure support. The interesting feature of ad hoc networks is the absence of base stations. Conventionally, base stations act as repeaters, relaying signals from one mobile station to another. In ad hoc networks, however, each mobile station is specially designed with the capability to relay signals for peer-to-peer communication. Thus, virtual paths are established as the need arises.

Since no base station is needed, ad hoc networks can be employed easily and rapidly. This factor makes the concept attractive for communications in situations, where the instrastructure is not available. Consequently, the cost of implementation is greatly reduced. In addition, it is envisaged that the network capacity will increase on par with the rising demand for mobile communications.

This research aims at investigating the different aspects of networking. The teletraffic performance of the network will be examined through computer simulation.

Figure 2 : An example of ad hoc network with two virtual paths, one links mobile stations B, J, I, E, A and G, while the other inks mobile stations C, E, H and D.

 

Mobile social networking aided content dissemination in heterogeneous networks
Prof. Lajos Hanzo, Prof. Lie-Liang Yang, Dr. Jie Hu

[1] Distributed Cooperative Social Multicast Aided Content Dissemination in Random Mobile Networks

[2] Throughput and Delay Analysis of Wireless Multicast in Distributed Mobile Social Networks Based on Geographic Social Relationships
A distributed multistage cooperative-social-multicast-protocol-aided content dissemination scheme is proposed, which is based on a self-organized ad hoc network of mobile stations (MSs) seeking the same content. In our content dissemination scheme, upon receiving the content, the content owners may further multicast it to their social contacts who are hitherto unserved content seekers. Then, we mathematically define the geographic social strength to describe the social relationships between a pair of MSs. By jointly considering the geographic social strength, the geographic distances, and the path loss, as well as the small-scale fading, we derive the closed-form formula of the average social unicast throughput. Furthermore, we model the content dissemination process by a discrete-time pure-birth-based Markov chain and derive the closed-form expressions for the statistical properties of the content dissemination delay. The proposed multistage cooperative social multicast protocol is capable of successfully delivering the content of common interest to all MSs in two transmission frames, provided that the density of the MSs is sufficiently high, as demonstrated both by our simulation and analytical results.


Figure 1. An example of the social multicast. The content owner (CO) is only willing to multicast the content of common interest  to its social contacts. A successful content delivery from a transmitter to a receiver depends on the following two conditions: (1) The receiver is a social contact of the transmitter; (2) The signal-to-noise-ratio (SNR) at the receiver end has to exceed a pre-defined SNR threshold for the successful content reception. A pair of MUs share a social relationships with a specific probability, which is defined by their geographic social strength. The content can be successfully delivered by a wireless link connecting a transmitter and receiver pair with another specific probability, which is determined by the related physical layer model. By jointly considering both the social relationship and wireless propagation characteristics, we can transform the left-hand-side `Real Social Multicast Scenario' to the right-hand-side `Mathematically Equivalent Model', where a receiver is connected to a transmitter by a `social wireless link'.














Figure 2. Multi-stage cooperative social multicast protocol.  During Frame 1, the CO set {CO1, CO2} cooperatively multicasts the content to the unserved CS set {CS1, CS2, CS3}. At the end of Frame 1, CS1 successfully receives the content and joins the CO set as CO3. Thus, during Frame 2, the new CO set {CO1, CO2, CO3} cooperatively multicasts the content to the unserved CS set {CS2, CS3 }. By the end of Frame 2, CS2 successfully receives the content and joins the CO set as CO4. Finally, during Frame 3, the new CO set {CO1, CO2, CO3, CO4} carries out the last stage of cooperative social multicast, and successfully delivers the content to the last unserved CS3.

[3] Mobile Social Networking Aided Content Dissemination in Heterogeneous Networks

[4] Cooperative Multicast Aided Picocellular Hybrid Information Dissemination in Mobile Social Networks: Delay/Energy Evaluation and Relay Selection

[5] Bridging the Social and Wireless Networking Divide: Information Dissemination in Integrated Cellular and Opportunistic Networks


An increasing number of Mobile Users (MUs) share common interests in general information, such as traffic information, weather forecast, and domestic/international news, etc. However, the Centralised Infrastructure (CI) based system has not been designed for efficiently disseminating the Information of Common Interest (IoCI) to numerous requesters. Thanks to the rapid development of mobile devices equipped with large storage and multiple communication modes, opportunistic communication between a pair of MUs can be readily realised. With the aid of opportunistic networks formed by MUs, we can improve the connectivity of cellular networks in the rural areas, we can offload the tele-traffic from the overloaded cellular networks and we can efficiently disseminate the IoCI in the densely populated areas. We commence with a detailed survey on the cross-disciplinary research area of Social Network Analysis (SNA) aided telecommunication networking. We continue by focusing our attention on the family of integrated cellular and large-scale opportunistic networks, whose performance is dominated by the inter-contact duration as well as the contact duration between any pairs of MUs. A Continuous-Time-Pure-Birth Markov Chain (CT-PBMC) is invoked for analysing the relevant performance. We demonstrate that the delivery ratio of the IoCI before it expires becomes higher than 90% with the aid of opportunistic networks consisting of 20 MUs. Moreover, our experiments based on the InfoCom 2006 mobility trace show that the opportunistic networks are capable of offloading 58% of the tele-traffic from the cellular networks. Thereafter, we propose a hybrid information dissemination scheme for the integrated cellular and small-scale opportunistic networks, which is comprised of two main stages, namely the BS-aided single-hop multicast stage and the cooperative multicast aided spontaneous dissemination stage. In contrast to their large-scale counterparts, in small scale opportunistic networks, the information dissemination is mainly affected by the mobility of the MUs, by the wireless channel attenuation and by the resource scheduling, where a Discrete-Time-Pure-Birth Markov Chain (DT-PBMC) is utilised for characterising the relevant performance. We demonstrate that our hybrid information dissemination scheme outperforms the traditional Base Station-aided Single-Hop Multicast (BSSHM) in terms of various delay metrics, despite consuming less energy.





















Figure 3. Two application scenarios for the integrated cellular and large-scale opportunistic networks. (a) Extending the coverage of cellular networks. (b) Off-loading tele-traffic from cellular networks. MUs may fetch the information of common interest from the BS when they enter the transmission range of the BS. Afterwards, they may disseminate the IoCI to their unserved peers who enter the transmission range of the information owners (IOs).  Epidemic protocol based on the store-carry-forward methodology is invoked for disseminating the IoCI amongst MUs.



Figure 4. Hybrid information dissemination scheme in a integrated cellular and small-scale opportunistic network. The hybrid information dissemination scheme consists of two main stages, namely (1) the BS-aided multicast stage and (2) the cooperative-multicast-aided spontaneous dissemination stage. Frame structures for these two stages are portrayed in this figure. Apart from the time slots (TSs) reserved for control signalling exchange, the BS only takes a singel TS for multicasting the IoCI, while n_0 TSs are provided for allowing multiple information multicasters to cooperatively multicast the IoCI without being affected by the adverse effect of interferences and collisions.

 

 



Video 1. A brief introduction of the social network analysis aided information dissemination design in integrated cellular and opportunistic networks. The full-length demos for the information dissemination in both the integrated cellular and large-scale opportunistic network and its small-scale counterpart are provided at the end of this video. Please Enjoy it~

 

 

[top] [Main Page]

University of Southampton: 
Interests
Maintained by Michael Ng and Li Li | © 2006 University of Southampton