The very short pulses used in ultra-wideband (UWB) radio, often less than a nanosecond in duration, result in the receiver being able to resolve the UWB signal's time-of-arrival (ToA) with very fine resolution.  This enables potential applications in high-resolution ranging, applications as varied as search and rescue and warehouse inventory control.

To achieve that goal this paper introduces a ToA measurement algorithm using generalized maximum-likelihood estimation, employing an iterative nonlinear programming technique to reduce complexity.  In addition, hundreds of blocked line-of-sight observations were analyzed to determine the probability distributions of the amplitude and time-of-arrival of an UWB signal over the direct path. These are used to provide a set of equations that determine the probability of over- and under-estimating range.

In verification experiments using the algorithm, it was found that the previously neglected effect of slowing of the speed of propagation that is caused by line-of-sight blockages resulted in significant excess delay, particularly at long distances, and hence in consistent overestimation of range by up to 5%.

• J.-Y. Lee and R.A. Scholtz, "Ranging in a Dense Multipath Environment Using an UWB Radio Link," IEEE Journal on Selected Areas in Communications, Vol. 20, Dec 2002, pp. 1677-1683.
• M.Z. Win and R.A. Scholtz, "Energy Capture versus Correlator Resources in Ultra-wide Bandwidth Indoor Wireless Communications Channels", Proc. MILCOM, Vol. 3, Nov. 1997, pp. 1277-1281.
  
• J.M. Cramer, R.A. Scholtz and M.Z. Win, "Evaluation of an Ultra-Wideband Propagation Channel," IEEE Transactions on Antennas and Propagation, Vol. 50, May 2002, pp. 561-570.  (This paper received the 2003 A. Shelkunoff Transactions Prize Paper Award of the IEEE Antennas and Propagation Society.)

A soft-inverse for a system is the locally optimal processor that updates soft (decision) information on the digital input and output variables. For example, a turbo decoder uses a soft-inverse for each of two constituent decoders and iteratively activates these soft-inverses or soft-in/soft-out (SISO) decoders.

A particularly important type of system is a finite state machine (FSM), which models the constituent codes of a turbo code, as well as many other channels and encoders in digital communication systems. The standard method for implementing the soft-inverse of a finite state machine (FSM) is the forward-backward algorithm (FBA). The FBA may be viewed as a generalization of the Viterbi algorithm - i.e., the FBA runs one Viterbi-like recursion on the FSM trellis in each the forward and backward directions. As such, the FBA inherits the so-called "ACS (add, compare, select)-bottleneck" that makes fast implementation of the decoder difficult.

In this work, we found a tree-structured architecture for computing the soft-inverse of an FSM. Compared to the FBA, this provides an exponential speed-up by avoiding the classic ACS-bottleneck. We arrived at this architecture by showing that the soft-inverse problem is equivalent to another simple computational problem: computing partial sums. Thus, the "tree-SISO" developed is essentially the SISO version of a fast, tree-adder.

• P. A. Beerel and K. M. Chugg, "A Low Latency SISO with Applications to Broadband Turbo Decoding," IEEE J. Selected Areas in Communications, vol. 19, May 2001, pp. 860-870.
• P. Thiennviboon and K. M. Chugg, "A Low-Latency SISO via Message Passing on a Binary Tree," Allerton Conf., Urbana, IL, Oct. 2000.
• G. Fettweis and H. Meyr. "Parallel Viterbi algorithm implementation: Breaking the ACS-bottleneck," IEEE Trans. Commununication, 37:785-790, August 1989.
• K. M. Chugg, A. Anastasopoulous, and X. Chen, Iterative Detection: Adaptivity, Complexity Reduction, and Complexity Reduction, Kluwer Academic Publishers, 2000.