 |
 Wideband CDMA
TDMA Architectures
Image and Video Compression
Link Layer Error Characteristics
Coding and Modulation
|
| A) |
Wideband CDMA
This collection of projects is motivated by several well known attributes of spread-spectrum waveforms, namely their ability to efficiently make use of the multipath on the channel (providing the spread bandwidth of the signals exceeds the coherence bandwidth of the channel), to allow high data rate multimedia traffic, permit spectral overlay of narrowband waveforms, and maximize the privacy of the transmitted signals. The research itself, briefly described below, consists of both terrestrial and satellite components, and further consists of scenarios relevant to both commercial and military communications.
|
|
i) |
Multicarrier CDMA
The use of multicarrier (MC) DS CDMA waveforms offers several advantages over the conventional single carrier system. With appropriate signal processing at the receiver, MC waveforms can achieve the equivalent of RAKE reception of a single carrier system. However, MC signaling does not require a continuous spectrum, and can be effectively used in an overlay situation without the need for adaptive notch filtering. The current research we are engaged in combines the multicarrier approach with forward error correction coding, in the sense that with M carriers, a rate 1/M convolutional code is employed. This project will continue through 4Q97.
|
|
ii) |
Performance of CDMA over Satellite Links
Our research on the use of satellites for mobile communications has primarily been concerned with performance analysis. Because of both the large round trip time delay inherent in satellite-based systems, and the relatively small multipath delay spread, CDMA systems perform significantly worse than they do on typical terrestrial links. We have quantified the effect of power control error in mobile satellite systems, and have recently extended the results to include the use of unmanned airborne vehicles (UAV). This project will extend through 4Q97.
|
|
iii) |
Interference Suppression in CDMA
This research has two components. The first component is designed for narrowband interference rejection, and has as its applications both spectral overlay of conventional waveforms and enhanced anti-jam capability for military scenarios. The second component is related to suppression of multiple access interference: The approach we have taken is the use of adaptive signal processing for interference suppression, as opposed to interference cancellation, because of the need in the latter schemes for accurate parameter estimation (such as received amplitudes or phases). The expected completion date is 2Q97.
|
|
iv) |
CDMA with Higher Order Alphabet Sizes
In order to allow very large data rates over a mobile channel, the use of higher order signal constellations is being considered. This work is relatively new, and is related to the CWC project on wireless LANs. A goal is the design of a system by 4Q96.
|
|
v) |
Acquisition Considerations in CDMA
We have done fairly extensive research on this topic, including designing both optimal and suboptimal receiver structures, deriving performance results, and incorporating antenna diversity into the system. In addition, we have looked at the effect of narrowband overlays on the acquisition performance. The expected completion date is 3Q96.
|
|
vi) |
Adaptive Spatial Arrays for CDMA
This research project is just beginning. The intent is to employ adaptive beamforming networks to enhance the capacity of cellular-type systems. This work will continue through 4Q99.
|
|
vii) |
Design of Broadband CDMA System for Wireless Multimedia Networks for the Digital Battlefield
This is a large scale research effort involving many researchers and various industrial collaborators. The intent is to design and evaluate the performance of a broadband CDMA system capable of supporting high data rate, multimedia, traffic, in the presence of battlefield conditions (i.e., intentional jamming and physical destruction of resources). Further, the system is designed for total mobility, meaning mobility of both handsets and base stations. Research is currently underway at all layers of the conventional network protocol stack.
|
|
viii) |
Use of Wideband CDMA for Naval Applications
This is a program being done jointly with researchers from NRaD; its goals are the design of new radios capable of supporting both ship-to-ship and ship-to-shore secure communications in a battlefield environment. The system is designed around the use of both satellites and UAVs to serve as either relays or intelligent base stations. The scope of the research is quite broad, incorporating wideband spreading, adaptive notch filtering, and adaptive spatial arrays. This work will continue through 2Q98.
|
[top of page]
|
| B) |
TDMA Architectures
|
|
i) |
Continuous Phase Modulation (CPM) for Wireless Applications
A particular form of binary CPM is being studied, and its performance in the presence of both multipath fading and co-channel interference will be compared against that of GMSK, the modulation format being used in GSM. This work will continue through 3Q98.
|
|
ii) |
A TDMA-based, high data rate, wireless LAN, will be designed and its performance analyzed. Effort shall extend over a two year time interval, beginning in July 1996. This project is funded from the CenterÕs budget.
|
[top of page]
|
| C) |
Image and Video Compression for Wireless Communications
|
|
i) |
Unequal Error Protection for Compressed Bit Streams
Image and video compression algorithms typically produce a bit stream in which the bits have different levels of importance. For example, there may be a header which gives information about image size, quantization tables, and other parameters essential for reconstructing the compressed images. As another example, in a hierarchical compression scheme, there may be portions of the data stream which refer to low resolution versions of the image, and less important portions which refer to successively finer detail. In these and other situations, transmission of compressed images and video on a noisy communications channel might be enhanced by providing higher degrees of error protection to the more significant portions of the compressed stream, at the expense of reduced protection of less significant portions. We propose to develop and evaluate methods for non-uniform allocation of available error protection bits among the different portions of the compressed stream. This project will continue through 4Q99. This project is funded from the Center's budget.
|
|
ii) |
Error Control Methods for Inter-Frame Video Compression
Intra-frame coding provides a simple way of controlling the impact of errors when transmitting compressed video over a noisy wireless communication channel. In this approach, each video frame is compressed separately and independently from other frames, so errors in the transmission of a single compressed frame influence the reconstruction of that frame only. Inter-frame coding, which uses prediction and interpolation methods to take into account the frame context, can achieve substantially higher levels of compression than purely intra-frame coding, but is vulnerable to error propagation. However, some of the compression gains from inter-frame coding might still be realizable in the setting of a noisy channel by appropriate application of efficient error control coding techniques. We propose to investigate, for a variety of compression schemes, the potential benefits of combining inter-frame prediction and interpolation with error correction coding to achieve reliable, low-rate transmission of compressed data over a noisy wireless channel. This project will continue through 4Q99.
|
[top of page]
|
| D) |
Link Layer Error Characteristics
Link-layer performance clearly depends on the error characteristics of the underlying physical network. Traditionally, the bit error rate is used as a measure of "channel quality". However, the dynamic behavior of packet based communications requires a more thorough understanding of the error characteristics. When considering the trade-offs involved in wireless link/physical layer design, it is important to study more dynamic measures of physical layer communications performance such as channel error correlation, burst error length, and block error probability.
Power control, interleaving and coding are traditionally used to compensate for burst errors caused by fading. These techniques conceal the bursty nature of the channel at the cost of interleaving and coding delay, plus additional system complexity. For packet based communications over bursty error channels, "masking" the underlying channel burstiness may simplify protocol design but packet delay and system complexity may suffer. Using the channel intelligently (e.g., hybrid ARQ, channel sensing, burst transmission), reasonable throughput may be achieved with better delay performance and lower complexity. In designing intelligent burst channel protocols, knowledge of the error characteristics of the underlying dynamic channel is necessary.
We shall also investigate the block error statistics that result when transmissions occur over a channel that is characterized only by its second order statistics with no further assumptions about its structure. Bounds on the block error probability are being derived. Analysis and simulation will also be performed.
|
[top of page]
|
| E) |
Coding and Modulation
Several research projects in the area of error-control coding, modulation codes, and spectrally-efficient coded-modulation techniques for wireless communication channels shall be initiated.
|
|
i) |
"Turbo" coding is a recently introduced technique that has generated considerable interest as a result of its ability to achieve reliable data transmission at signal-to-noise ratios close to Shannon's bound for a specified code rate. The components of a turbo coding system are a parallel concatenated convolutional encoder with permutation ("interleaving") of the encoder inputs and an iterative, "soft-output" decoding algorithm. There are many basic issues related to turbo code design that require further investigation, such as: selection and structure of component convolutional codes, size and selection of the input permutations ("interleavers"), and improved theoretical foundations for analysis of code weight distributions and other structural properties. Wireless communication channels characterized by multipath fading and inter-user interference, in addition to additive white gaussian noise (AWGN), will likely alter the code design criteria that have begun to emerge from studies of their behavior on AWGN channels. Commercial personal communication systems will also impose practical constraints on acceptable coding technology that will impact the design criteria that turbo codes (and other forward-error-correction codes) must satisfy. This project is funded from the Center's budget.
|
|
We propose to develop and analyze coding methods based upon turbo coding principles for possible use in wireless applications. This project will continue through 4Q99.
|
|
ii) |
Trellis-Coded-Modulation (TCM) techniques have revolutionized data transmission over bandwidth-limited channels during the past decade. The availability of 28.8 kbps modems (such as the one in the PCMCIA card slot of the notebook computer used to prepare this project description) for voice-band data communications gives testimony to the dramatic technological impact of the spectrally efficient coded- modulation techniques introduced by Ungerboeck in the early 1980Õs, and extended and refined during the following decade. The application of TCM to channels with other characteristics - particularly fading channels - and with additional constraints - such as reduced-complexity decoding for high data rate applications - provides some foundation for the investigation of TCM in wireless communication systems. As in the case of turbo codes, however, the time-varying nature of the wireless channel, and the co-channel interference inherent in multi-user systems, demand novel approaches to designing efficient and practical TCM schemes for this application.
|
|
We propose to assess the strengths and weaknesses of existing TCM methods and explore new techniques for coded-modulation that can enhance the performance and data transfer capabilities of various wireless communication systems. This project will continue through 4Q99.
|
|
iii) |
Constrained codes - also known as modulation codes - play an important role in many digital data transmission and recording systems. By endowing the channel input signals with various time-domain properties (such as runlength limitations) and frequency-domain properties (such as bandwidth compression and spectral null frequencies), these codes enhance the reliability of signal recovery by reflecting characteristics of the recording process, channel transfer function, detection method, control loops (timing recovery, gain control, adaptive equalization), and electronics that are often not explicitly considered by other coding techniques such as those discussed above. It has also been demonstrated that certain modulation codes have signal distance- enhancing characteristics that make them attractive in coded-modulation schemes.
|
|
Techniques for constrained code analysis, design, and application to coded-modulation have experienced major progress in recent years, and the relevance of these advances to improving transmitted signal characteristics in wireless communication systems has not received much attention to date.
We propose to study the requirements for constrained code design in wireless data transmission systems and to develop and apply, as appropriate, the improved code construction methods to generate practical, efficient modulation codes. This project will continue through 4Q99.
[top of page]
|
|
|
|
 |
|
Affiliated Faculty
Research Thrust Areas