Jeffrey M. Capone - US grants

In integrated services wireless networks, geographically distributed heterogeneous applications must share a common wireless channel. To coordinate the sharing of the transmission resources, a multiple access protocol such as time-division multiple access (TDMA) or code-division multiple access (CDMA) is required. To control QoS in an integrated services network, usually both call admission and resource management (scheduling at the packet level) among the admitted calls are needed.

At call admission, it must first be determined if there exists a slot allocation (as in the case of TDMA) or power management (as in case of CDMA) policy that can deliver a target QoS to each application in the
network. To answer this question, the region of achievable QoS (set of performance vectors that can be delivered under any admissible policy) needs to be identified. The region of achievable QoS will depend on the underlying channel conditions and the multiple access technology. Each access technology has its inherent advantages and disadvantages in combating the effects of the wireless channel and in its ease of accommodating variable rate service while delivering diverse QoS requirements. The determination of the region of achievable QoS leads to the development of call admission and scheduling functions.

The PI has been conducting research in the area of integrated services over wireless networks for the past five years and now proposes to translate some of his and others prior research into viable technology
and educational tools through the development of an integrated services wireless network (ISWN) test-bed. The ISWN project is multifaceted in the sense that it contains educational, development, and research components. The project will be implemented in three phases: system design and specification, development of novel approaches to support multi-media traffic and software implementation in an operational system.

9979403
Balanis

The antenna is one of the fundamental distinctions between a wired and a wireless system. The design of the antenna impacts the development of each component - from the circuit design to the receiver structure and coding technique, as well as the channel access protocol - employed in future wireless communication networks. The most challenging environment for the design of each component is where the communication devices are extremely small, low power computing and consuming mobile terminals. Furthermore, these terminals may be able to move randomly and organize themselves in an ad hoc communication structure. Such a network of communication terminals is referred to as a mobile ad hoc network (MANET). It is proposed to investigate the use of adaptive antennas (i.e., smart antennas) to improve channel quality in a MANET. The design of a smart antenna for a MANET will pose many challenges in antenna, feed network, signal processing, receiver, and protocol designs.

The antenna must be designed to be conformal to the portable device and to adapt to the needs of the MANET. Depending on the needs of the MANET, the antenna will be designed to have different modes of operation from omni-directional to highly directive. A single antenna element will not be able to achieve these characteristics and therefore, a collection of elements, referred to as an array, will be needed. The array is a versatile antenna system and can perform many different functions (including scanning, beam forming, etc.) by primarily controlling the excitation amplitude and phase of its elements. The feed network devices and circuitry, and the signal processing algorithms which must be developed and also integrated into the entire system, will provide the desired excitation of the elements for the antenna to meet the needs of the MANET. The challenges to achieve these objectives include selection of the antenna element and array geometry, and the modeling of the antenna system. The physical realization of the selected antenna configuration and the integration of the antenna and feed network models as well as the analysis of the effect of the antenna on the overall system performance will be investigated. In addition, the modeling of the propagation channel (multi-path, fading, etc.) will also have fundamental impacts on the overall antenna system design.

The antenna feed network, consisting of controlling and matching devices and circuits, will require extensive analysis and design tools. To accomplish the tasks presented in this proposal, a large number of circuits operating at future communication frequencies (20-GHz range), including matching networks, low-noise and power amplifiers, oscillators, phase shifters, and mixers must be developed. The circuits will be highly compacted and fabricated in several layers to fit within the relatively small area available in a communication device. This task requires clever selection of the semiconductor material for fabrication, optimized semiconductor device structures, appropriate high-frequency circuit designs, and highly accurate device and circuit analysis tools. Moreover, the signal processing algorithms used in driving these circuits must be considered in the overall design.

The signal processing algorithms, to control and adapt the different modes of operation of the antenna, will be developed for analog and digital antenna beam-forming at high (20 GHz) frequencies. This work will include the development of algorithms for computation of gains and phases for use in analog phase shifters for the antenna array. Although phase shifters will be initially realized in the analog domain, a DSP algorithm will compute the gain and phase. The challenge here lies in the frequent computation of antenna parameters intended for operation in a rapidly changing environment such as a MANET. Digital realizations will be addressed by developing spatio-temporal signal processing algorithms for beam-forming. Channel coding may als0 be employed to provide diversity and to suppress residual interference for multi-user communications over fading channels. By using channel coding techniques, the same error rates with a lower signal-to-noise ratio may be obtained resulting in an improved battery life of each mobile terminal.

From a networking point of view, smart antennas may improve the performance of current channel access techniques for MANETs and may allow for the development of new protocols. By employing an array of antennas, it may be possible to develop protocols to control the multiple access interference in the spatial dimension, as opposed to time or frequency, to further improve the performance delivered to the terminals operating in a MANET. The design of such channel access protocols is dependent on the signal processing technique used in the smart antenna system.
***

The organizers, Sourav Bhattacharya (sourav@asu.edu), Jeffrey M. Capone (jcapone@asu.edu), and Kevin Dooley (kevin.dooley@asu.edu) will hold an exploratory workshop to investigate the research issues in quality and process aspects in the engineering and operation of complex network systems. The focus and motivation for this workshop stems from recognizing the immense practical utility and impact that software process and quality research has made to the development of software - leading to the subject area of "software engineering". The objective of this research is to create similar impact in the network engineering and operations area. In Network Engineering & Operations (NEO), design process and quality-oriented research is largely missing. The state-of-the-art in NEO process is similar to the unstructured approach of software development that is clearly outdated now, and often considered harmful. A good NEO process should allow a network developer or engineer to commence with a set of requirements, specify them, analyze them, consider high level design alternatives, adopt an implementation, operate the network efficiently, and test them in a cost effective, controlled and evolutionary manner.

The focus of this workshop would be to discuss the research issues in NEO. The researchers will address research issues both along the lines of "Underlying Network Technology" and "Network System Engineering" area. Examples of such research issues will include, but are not limited to, the following: QoS and QoS testing; end-to-end integration; commercial-off-the-shelf (COTS) components compatibility and interface; traffic moderation and large-scale flow control; tools for life-cycle management, configuration control and network resource management; management of network scalability and growth; design and testing for availability, assurance, and cost-effective maintenance. The workshop will foster active participation from academia, industry and government. It will include invited key researchers from core network technologies area, as well as systems engineering, network operations and process & quality areas. The basic nature of this research is highly interdisciplinary, calling upon experts in the area of computer science, electrical engineering and process & quality (manufacturing). Participation in this workshop is open to all interested parties and contributors. Outcome of this workshop will include identification of some of the research issues in NE, and a roadmap to carry forward to the next step.