Plenary Talks

Professor fred harris, PhD, FIEEE
Electrical & Computer Engineering
San Diego State University, USA

Speech Title: Next Generation Receivers: Coupled Perfect Reconstruction Analysis and Synthesis Filter Banks

Abstract:
A novel and surprisingly flexible and computationally efficient set of new signal processing options are to be found in systems that contain coupled down-sampling analysis filter banks and up-sampling synthesis filter banks on each side of a communication link. The coupled filter banks convert high bandwidth high sample rate signal streams to multiple parallel low bandwidth, low sample rate streams. The energy required to process the multiple low sample rate streams from the analysis filter bank is reduced by more than an order of magnitude compared to the energy required to process the single input, high sample rate, stream. The modified multiple, low sample rate, data streams are reassembled to form the modified high sample rate output stream. The signal modifications applied to the low sample rate signal set include matched filtering, equalization, timing recovery, carrier recovery, diversity combining, and beam forming. The reduced workload has a significant and desirable impact on battery life and processer costs in the physical layer of future wireless systems. In this talk we will show the evolution of the signal processing operations starting with the Armstrong architecture through equivalency transformations that lead to the M-path filter bank and then on to the coupled M-path analysis-synthesis filter banks.

Professor Kin K. Leung, PhD, FIEEE
Electrical and Electronic Engineering
Imperial College, London, UK

Speech Title: From Compressed Sensing to Distributed Signal Processing

Abstract:
In this talk, the speaker will give a brief overview of compressed or sparse sensing to illustrate the use of various performance measures for tracking and prediction. By considering sensor networks with sparse signals and temporal correlated samples, he will show how a hybrid-Bayesian-Kalman technique can be helpful in exploiting both the sparsity and time dynamics of signals. Numerical examples using NASA ozone measurements will be provided. As the second part of this talk, the speaker will move on from centralized to distributed signal processing where signals (data) are fused (aggregated) along the way they are transferred toward the end user in the sensor networks. It will be shown that finding the optimal solution for the distributed processing problem is NP-hard, but a constrained optimization problem can lead to a fully distributed solution framework where optimal decisions based on local information at each sensor node can yield the global optimal solution. Future work on integrating compressed sensing and other signal processing techniques with the distributed solution framework will be discussed.

Professor W.C. Siu, PhD, DIC, CEng, FHKIE, FIET, FIEEE
Electronic and Information Engineering
The Hong Kong Polytechnic University, HK

Speech Title: Transform Domain Processing for Recent Signal and Video Applications

Abstract:
Orthogonality and Cyclic convolution Structure are two major properties of many transforms to be able to completely recover a signal from its version in the transform domain. In this talk we start with a visit on some very fundamental requirements of a transform with these properties and naturally formulate the discrete Fourier transform (DFT) and the number theoretic transform (NTT). We highlight the fastest way to realize DFT using NTT, and then talk about the discrete cosine transform (DCT) and subsequently the newest development of the integer cosine transform with the kernel (1,2,1) in H.264 standard and our newly suggested kernel (5,7,3) for video coding. We shall also discuss our recent work on transcoding in the DCT domain, and the newest results on using DCT techniques for fast and quality video interpolation. We have a strategy to use our academic research results to underpin industrial research, which is mostly related to modern video surveillance with big data and IoT (Internet of Things). Fruitful demonstrations and illustrations will be included, and the presentation will end with brief ideas on new trends and future applications.