(Color online) SIC with embedded ATR and MP for a two user system. Each user is decoded in succession, with interference removed with each iteration. Estimated symbols are combined with channel updates to estimate the MAI. SIC applies the proper Doppler correction to the interference before synchronization, scaling, and subtraction.
(Color online) Illustration of processing a single block with three iterations of the proposed receiver for a system with two users. Dashed lines represent channel estimates from the previous block (or estimated during training) eventually replaced with channel estimates from the current block. Notional representations of constellations and impulse responses are shown below the block being processed.
Block diagram of ATR processing targeting a user k. The filter weights are designed to balance matching the channel impulse response from user k and minimization of MAI. Following combining, a short DFE is employed to combat any residual ISI present in the effective channel.
(Color online) Reconstruction of a portion of the KAM11 experiment in 100-m-deep water. A moving source was towed at 35 m depth at a radial velocity of approximately 1.3 m/s (R = 1.3 km). The data example combined the moving source transmission at 1.3 km range (circle) and the stationary source transmission at 3 km range (square) to the receiving array. The second deepest hydrophone marked by the open circle was malfunctioning during this portion of the experiment. Data from this hydrophone were not considered in this analysis.
(Color online) Analysis of time-varying channel for (a) stationary, and (b) moving sources after Doppler compensation: (Top) Initial CIRs between both sources and each array element estimated in the presence of MAI; (middle) CIR estimates between both sources and each array element estimated after MAI has been removed after multiple iterations of the SIC process; and (bottom) time-varying CIRs between both users and a single receiver element at 79-m depth estimated with the MP algorithm after interference cancellation during iterative processing.
Scatter plots of soft symbol estimates of moving and stationary users after each of four iterations of the receiver. Decoded symbols are fed back to improve the interference cancellation, aiding the receiver in future processing iterations.
(Color online) (a) Aggregate symbol error rate for each user after each iteration of the interference cancellation receiver; and (b) performance comparison between processing of the packets prior to combining with a single-user receiver (solid curves) to multiuser processing of the combined packet with the proposed multiuser receiver (dashed curves).
(Color online) (a) Histogram of SIC efficiency for the mobile user after the first and second iterations, and (b) cumulative distribution of SIC efficiency for each user after each iteration and over all 190 decoding blocks. Positive values represent successful interference removal with 1 dB interpreted as approximately 20% of interference removed as a percentage of total signal power.
Article metrics loading...
Full text loading...