Time and space domain filtering for improved HF communication

High speed data transfer has until recently been generally unattainable over HF channels because of the highly disturbed nature of the channel. The development of real-time signal processors has done much to mitigate effects of channel disturbances, making it possible to substantially increase throughput. Data rates as high as 9600 bits per second can be achieved by implementing signal processors as adaptive channel equalizers. This recent increase in data rates has rendered feasible HF spread spectrum communication systems. Spread spectrum systems are attractive in many applications because they are resistant to narrowband interference. However, channel equalizers necessary for high data rates are susceptible to narrowband interference. Spatial or temporal prefiltering has commonly been employed to prewhiten the received data, providing improved equalizer performance. The more popular one-dimensional techniques for temporal and spatial prewhitening of HF communication signals are reviewed. Also, combined space-time prewhitening techniques are proposed. Algorithms for designing and implementing these two-dimensional whitening filters are presented. Advantages and disadvantages of the two design strategies are discussed. Particular attention is focused on system performance, computational requirements, and cost. Computer simulated results for these signal processing algorithms are presented.