Development of Synchronization Techniques for Low-Power Orthogonal Frequency Division Multiplexing Applications, Project 14-R8030Printer Friendly Version
Inclusive Dates: 01/12/09 10/30/09
Background - Orthogonal Frequency Division Multiplexing (OFDM), a modulation scheme that has been in use by high-end specialized systems for several years, is migrating to consumer products. Its ability to operate in multipath and fading environments makes it an ideal choice for mobile wireless applications requiring very high data rates. However, like any communications scheme, OFDM has limitations, primarily, sensitivity to timing synchronization and intolerance of frequency offsets. To take full advantage of the potential benefits and capabilities of OFDM for low-power applications, SwRI initiated an effort to address these limitations.
Approach - The objective was to reduce risk in future product development by deriving solutions applicable to low-power configurations, developing the necessary algorithms, and implementing a receiver-channel thread using those algorithms. SwRI began by developing a baseline implementation using published synchronization methods/techniques. Matlab® code was generated that implemented OFDM signal generation and decoding. Over time, propagation delays, multipath, and frequency errors were added to the Matlab® implementation to simulate a real world environment and propagation path. Numerous analyses were conducted to evaluate the performance of these algorithms and determine their effectiveness. A hybrid Park/Schmidl approach was developed as a practical approach for low power systems. Significant testing and Monte Carlo-type simulations proved this approach to be robust, efficient and effective, even in the presence of significant signal-to-noise-ratio (SNR) degradation and multipath, in producing reasonable bit error rates (BER).
Once the approach was developed and simulated, the effort was refocused on validating the simulation results with real world experiments. A test bed system was developed using standard laboratory test equipment and components. All tests completed to this stage indicated that the developed algorithm worked successfully. The final effort of the project was to analyze the viability of incorporating the synchronization scheme into an actual digital signal processing (DSP) engine.
Accomplishments - The team was able to implement the approaches in a simulated OFDM link in Matlab®, as well as an actual one-way OFDM wireless link using laboratory equipment and development hardware. The data and knowledge obtained from both of these implementations has provided SwRI with added capabilities in the area of very high speed communication that was previously limited to conventional low-speed data links relying on serial data transmission techniques.