UCSC Broadband Communications  (UCBC)  Research Group   


Ad-hoc Networks

There are many different scenarios for wireless networks. In a traditional cellular wireless network (or wireless LAN), the nodes (or users) communicate with each other through base stations (or access points). They provide access control in the network, such as frequency use, power, coding, etc., playing a central role. The base stations are usually fixed and can form a network infrastructure while the nodes are either stationary or mobile. An example of such network is the common cellular phone system.

However, in situations where there is no fixed infrastructure, such as battle fields or  catastrophe control, wireless ad-hoc network is a better choice than the fixed cellular LANs. Thus, on these cases, there is no base station and all the control and access tasks are distributed among nodes. The lack of a centralized control system imposes real challenges for ad-hoc network designers and the research area has still a lot of unanswered questions like throughput-delay trade-offs and capacity bounds. There are also a lot of network modeling issues need to resolve that will ultimatley help better analysis of these networks.

We work on many different aspects of mobile ad-hoc wireless networks (MANET) related to packet delivery and the tradeoff between throughput and delay. For example, we have developed a novel multiuser diversity strategy for packet relaying, which permits more than one-copy (multi-copies) of a packet being received by relay nodes, thus allowing to decrease the delay on such networks for a fixed number of total users n.

Delay

d - delay measured when source node relay only one copy of a packet (K=1).

dk - delay measured when source node relays multiple copies of a packet (K>1).








Simulation results for the Random Waypoint Mobility Model. Each grey point is a pair (d,dk) delay measured for 40 random topologies all plotted together. A 7th degree polynomial fit for all the points and a 90 consecutive points average are plotted for K=2. The theoretical curve for the steady-state uniform distribution is also plotted.  Simulations parameters used were: n=1000 nodes, v=0.13m/s, r=0.02m, simulation area = 1 m^2, total simulation time = 1000s.



Interference

Also, we have analyzed interference effects for a large number of nodes n in the network cell, and shown that for the case where the path loss parameter alpha is greater than two, the resultant signal-to-interference ratio, for a receiver node communicating with a close neighbor, tends to a constant as n scales to infinity, regardless of the position of the receiver node in the cell. Therefore,  communication is feasible for close neighbors when number of interferers scale to infinity. For the receiver nodes at the boundary of the circular cell, we show that they suffer less interference than those inside.










Signal-to-Interference Ratio curves as a function of n, for 3  <= alpha <= 6, and the receiver node considered located at the center (r'=0) and at the boundary of the cell (r'=1/pi^{1/2} - r_o).