This PDF file contains the front matter associated with SPIE Proceedings Volume 6577, including the Title Page, Copyright information, Table of Contents, and the Conference Committee listing.

This paper presents a concept for a United States Coastal Area Network (U-SCAN) that is comprised of IEEE 802.11, 802.16, and satellite communications technologies. The Office of Naval Research (ONR) on behalf of the National Oceanographic Partnership Program (NOPP) has tasked The Johns Hopkins University Applied Physics Laboratory (JHU/APL) to perform an architectural study into the establishment of a United States Coastal Area Network (U-SCAN). The goal of this study is to define a wireless network architecture that can be deployed to enable contiguous coastal area network coverage for scientific, commercial, and homeland security (e.g. Coast Guard) applications within the United States Exclusive Economic Zone (EEZ), in a manner that is flexible, manageable, and affordable. The JHU/APL study will ultimately provide recommendations to NOPP regarding potential network architectures and technologies that could provide the desired capability, with a particular focus on commercial (both existing and emerging) technologies. This paper presents the envisioned U-SCAN architecture, and presents the envisioned technical capabilities and shortcomings of the component candidate technologies.

We consider the problem of signature waveform design for code division medium-access-control (MAC) of wireless sensor networks (WSN). In contract to conventional randomly chosen orthogonal codes, an adaptive signature design strategy is developed under the maximum pre-detection SINR (signal to interference plus noise ratio) criterion. The proposed algorithm utilizes slowest descent cords of the optimization surface to move toward the optimum solution and exhibits, upon eigenvector decomposition, linear computational complexity with respect to signature length. Numerical and simulation studies demonstrate the performance of the proposed method and offer comparisons with conventional signature code sets.

Mobile ad hoc network (MANET) is a radio packet network without dedicated infrastructures. In recent years it has received tremendous attention because of its self-configuration and self-maintenance capabilities. However, because of node mobility and shared wireless links, its routing protocol design presents nontrivial challenges such as broadcast storm, stale route and delay. This paper proposes a location-based route self-recovery technique for source-initiated routing protocols. The purpose of route self-recovery is to reduce overhead and delay during route maintenance as well as allowing continuous packet forwarding for fault resilience. The ns-2 based simulation shows throughput and overhead improvements of source-initiated routing with route self-recovery and in the case of highly dynamic environments and heavy traffic loads, it is more robust and scalable than other protocols.

Ad hoc networks rely on cooperation in order to operate, but in a resource constrained environment not all nodes behave altruistically. Selfish nodes preserve their own resources and do not forward packets not in their own self interest. These nodes degrade the performance of the network, but judicious route selection can help maintain performance despite this behavior. Many route selection algorithms place importance on shortness of the route rather than its reliability. We introduce a light-weight route selection algorithm that uses past behavior to judge the quality of a route rather than solely on the length of the route. It draws information from the underlying routing layer at no extra cost and selects routes with a simple algorithm. This technique maintains this data in a small table, which does not place a high cost on memory. History-based route selection's minimalism suits the needs the portable wireless devices and is easy to implement. We implemented our algorithm and tested it in the ns2 environment. Our simulation results show that history-based route selection achieves higher packet delivery and improved stability than its length-based counterpart.

Source localization is investigated based on noisy measurements of TDOA (time-difference of arrival) in which the measurement noises are assumed to be Gauss distributed. The solution to the constrained WLS (weighted leastsquares) is derived and applied to the source localization problem based on TDOAs. The proposed algorithm is shown to be an approximate MLE (maximum likelihood estimation) algorithm under some mild condition. The simulation results show that the proposed approximate MLE algorithm compares favorably with the existing solution methods for source localization based on TDOA measurements.

Localization and tracking of the ground moving target (GMT) are investigated based on measurements of TDOA (time-difference of arrival) and AOA (angle of arrival) in which the measurement noises are assumed to be uncorrelated and Gaussian distributed. An approximate MMSE algorithm is proposed via developing constrained Kalman filtering based on the pseudo-measurement model in the existing literature that leads to a nonlinear constraint imposed on the state vector for the GMT model. Randomization of the state vector suggests to replace the hard constraint by its expectation. We first derive a solution to a similar constrained MMSE problem that is used to extend the Kalman filtering to develop a linear recursive MMSE estimator subject to the nonlinear constraint as mentioned earlier which is termed as constrained Kalman filtering.

Most military applications of GPS require performing both jammer nulling and localization. While nulling can be achieved using adaptive gradient techniques applied to the input sampled data, high resolution direction of arrival estimation can precede using subspace and eigenstructure methods applied to the estimate of the covariance matrix. In this paper, we extract the jammer direction of arrival (DOA) information directly from the adaptive weights, in which case we assume interference nulling precedes interference localization. This high resolution DOA estimation approach based on available beamforming weight values leads to simplified receiver structure and allows a choice of IF or baseband processing as well as flexibility for analog, digital or mixed mode implementations.

A new spectral direction of arrival (DOA) estimation algorithm is proposed that can rapidly estimate the DOA of non-coherent as well as coherent incident signals. As such the algorithm is effective for DOA estimation in multi-path environments. The proposed method constructs a data model based on a Hermitian Toeplitz matrix whose rank is related to the DOA of incoming signals and is not affected if the incoming sources are highly correlated. The data is rearranged in such a way that extends the dimensionality of the noise space. Consequently, the signal and noise spaces can be estimated more accurately. The proposed method has several advantages over the well-known classical subspace algorithms such as MUSIC and ESPRIT, as well as the Matrix Pencil (MP) method. In particular, the proposed method is suitable for real-time applications since it does not require multiple snapshots in order to estimate the DOA's. Moreover, no forward/backward spatial smoothing of the covariance matrix is needed, resulting in reduced computational complexity. Finally, the proposed method can estimate the DOA of coherent sources. The simulation results verify that the proposed method outperforms the MUSIC, ESPRIT and Matrix Pencil algorithms.

Orthogonal Frequency Division Multiplexing (OFDM) is a widely used technique for data transmission on multipath fading channels. The multipath component of these types of channels causes a phenomenon known as frequency selective fading. This type of fading can severely degrade or completely eliminate the signal energy of many of the OFDM tones producing an irreducible error rate, even when no noise is present. In the early 1990's, researchers combined some of the characteristics of Code Division Multiple Access (CDMA) and Spread Spectrum (SS) with OFDM in order to create a more robust modulation scheme capable of surviving frequency selective fading without the need for forward error correction (FEC) techniques and thus OFDM-CDMA was born. Similar to OFDM, FEC coding can be added to OFDM-CDMA waveforms to further improve performance (referred to as COFDM-CDMA). An additional performance improvement possible for COFDM-CDMA is to exploit the output of the FEC decoder in an iterative fashion. This paper will investigate the possible performance benefits of feeding back the decoded FEC data bits to the mutiuser (i.e. multidata) detection (MUD) scheme of COFDM-CDMA waveforms on various HF multipath/fading channels.

This paper will investigate the use of an enhanced rate one Alamouti Space Frequency (SF) multiple antenna Orthogonal Frequency Division Multiplex (OFDM) radio communication system. A two transmit, single receive antenna system will be simulated to operate under conditions of multipath fading with noise. A simple modification to the standard coherent Alamouti receive combiner will be applied and shown to improve bit error rate (BER) performance on rapidly fading multipath HF channels. Orthogonal Frequency Division Multiplexing frequency domain techniques will be utilized to effectively eliminate the Inter-Symbol Interference (ISI) resulting from the effects of multipath. Numerically simulated results will be shown for several multipath fading High Frequency (HF) radio channels. Inner convolutional error correction coding will be applied in addition to the Alamouti coding and numerically simulated BER results presented. Various HF channel conditions will be simulated including the 2 ms, 10 Hz, 2 ms, 5 Hz, CCIR poor (2 ms, 1 Hz) and extra poor (2 ms, 2 Hz) channel conditions. Performance under conditions of correlated transmit antennas will also be investigated.

Since their introduction in 1993, turbo codes have received a significant amount of attention in the communications theory field due to their Shannon-capacity approaching performance. In recent years, the cellular systems market has embraced turbo code technology and made it part of the latest standards. This paper will review the effects of scale factors, fixed-point precision, soft decisions and hard decisions on the performance of the turbo codes defined in the UMTS third-generation cellular system. In addition, a new scale factor estimation technique which provides improved performance at low signal to noise ratios will be presented.

We present an integrated telemetry data link and network architecture system solution, developed by Mayflower for the Air Force, AFFTC, Edwards AFB, CA, based on third generation UMTS cellular standards. The data link, called COTS Affordable Data Link System (CADLS), accommodates high mobility user applications typical of tactical fighter aircraft. The data network, called Telemetry/TSPI Data Network (TDN), uses enhancements such as a multi-tiered network protocol structure to provide flexible IP-based transport, work with multiple air interface protocols, accommodate test platform mobility, and integrate seamlessly with unified infrastructure such as Test and Training Enabling Architecture (TENA). The end-to-end CADLS/TDN datalink network system, described in this paper, is a 2- way, asymmetric IP-based wireless network system, and as such it is a potential candidate technology to support the DoD CETIP integrated Network Enhanced Telemetry (iNET) project. The integrated CADLS/TDN system is at an advanced stage of prototype development. We present the integrated CADLS/TDN system architecture, its features and capabilities, and the laboratory prototype developed under the Air Force program.

Accumulative Predictive Error has been previously used for time series modeling of psychological response time data. In this paper we extend its applicabilty to the identification of tap amplitude statistics of ultrawideband communication channels. We also present channel modeling results from two other model selection techniques: Minimum Description Length and Akaike's Information Criterion. Channel models are also identified by hypothesis testing using Kolmogorov-Smirnov test. The results agree with recent findings that Rayleigh distribution can still be used to model tap amplitude statistics of line of sight ultrawideband communication channels.

This paper discusses recent studies on Chirp Slope Keying (CSK) as a scheme suitable for underwater communications and presents a new study on the performance of a time-frequency receiver in a Rayleigh fading environment. As expected, CSK proves to be a digital modulation scheme inferior in the AWGN channel as compared to traditional schemes but very promising in more detrimental channels present in underwater communications. In effect, while most schemes' performances decay abruptly with the addition of new disturbances, a time-frequency CSK receiver's performance deteriorates slowly with increasing Rayleigh fading. Intuition dictates that CSK will overpower other schemes as channel conditions continue to worsen. The receiver first uses the Wigner distribution to transform the incoming signal to the joint time-frequency domain and then computes the Radon transform to determine the binary digit received.

Digital fingerprinting is an available method that can identify the customer or owner of the digital media (image, video, etc) and protect the copyright of the content providers. Fingerprinting system embeds a unique fingerprint, which is the identity of the owner, into each individual copy of the original content. The monitor can detect the original user of the legal copies and track the usage of the copyrighted media content. The most serious problem to fingerprinting is how to resist the effective attack--collusion. Collusion is to use some different marked copies of the same original content and generate a new version from which the detector can't regain the correct fingerprint. In this paper we use combinatorial theory and construct nonlinear DBBD (Differential Balanced Block Design) code as fingerprint for digital media. The code is collusion-secure and has the capability of error correction which is beneficial to the robustness of the fingerprint. We also present a scheme for combining this collusion-secure fingerprinting code with a multi-resolution wavelet based watermarking mechanism. We use wavelet-watermarking techniques to embed and extract the fingerprints, and then use our detection scheme and the error correction ability of the DBBD code to find the pirates.

The conventional probabilistic argument has been used to analyze how many randomly distributed sensors are needed to cover a region. However, if these sensors move after deployment such as in mobile sensor networks, the probabilistic argument fails since the probability distribution is distorted. We provide an elementary argument that is not only as powerful as the probabilistic argument for analyzing the needed number of sensors to cover a predefined region, but also powerful enough to allow us to analyze the situations when the sensors move after deployment. This argument is the incompressibility method, using Kolmogorov complexity.