Description:
The overall research objective is to facilitate wireless global access to a broad range of communication services including voice telephony and access to the Internet. Toward this objective, spread spectrum CDMA transmission is now coming into service for cellular telephony and it is also under consideration for indoor microcellular telephony, CDMA is also being considered. These satellite systems aim to provide communication anywhere on the globe.
Fast synchronization is essential for satisfactory personal communications. The first objective of the proposed research is to improve the performance of PN code synchronization in dense user areas where the co-user noise is very high. Good synchronization is also a requirement for user capacity increase by interface cancellation. The proposed research is aimed at low earth orbiting (LEO) satellite systems which have large Doppler frequency shifts.
Sponsor Benefits:
The immediate benefit of this research is improved startup performance in CDMA personal communications. Rapid acquisition is important since personal communication systems require a startup time in the order of one second or faster.
People:
M.Sc. student: Renee Schmitz
M.Eng student: Derek Thille
Faculty members: David Dodds and Ralph Mason
Industry contact: Mike Moher(CRC)
Results to Date:
Based on theoretical study [1], the proposed research will use non-coherent combining of many short I&Q matched filters. This has a rapid decision rate and combines the Doppler shift tolerance of the short matched filter with the noise tolerance of the long matched filter. The performance is equivalent to the widely used traditional serial search with non-coherent detection except that the decision variable is available at the chip rate instead of the dwell time rate. The matched filter implementation is relatively simple and the result should provide a speed improvement in the order of 1000.
A system design (suitable for VLSI) has been developed for the receiver. The structure has been segmented to reduce the complexity of each chip. Several functional units can then be connected to produce the complete syncronizer and to allow testing on real signals. It is proposed to use the GPS signals for testing since they are well standardized and documented.
One segment of the VLSI structure has been developed as a VLSI class project by Ralph Mason at the Univ. of Regina for fabrication through the Canadian Microelectronics Corporation. Unfortunately there was a circuit level problem in the initially fabricated chips and recently a second layout has been submitted. The design is not complicated so we are reasonably confident of a working chip within a few months. Integration of the chips and system evaluation will be done starting in summer as part of an M.Sc. project.
[1] D.E. Dodds and M. Moher, 1995. "Spread Spectrum Synchronization for a LEO Personal Communications Satellite System", Proceedings of Canadian Conference of Electrical and Computer Engineering, pp. 20-23, Sept. 5-8, Montreal, Canada
Idusty Relevance:
A capacity of 40 or more active users per cell is desirable in order to reduce handoff between cells and to allow larger cells. Under these conditions, PN code acquisition and synchronization becomes a difficult problem. Syncronization is further complicated by multipath fading of the indoor channel and a doppler shift in the satellite channel. Synchronization in the presence of high noise will allow a high number of co-users which improves the system economics. Accurate syncronization is also a requisite for interface cancellation; a method to allow even more co-users in a cell.
This page last updated on July 23th, 1997.
Copyright © 1996 Ralph Mason.