Accessing remote data is a challenging task in mobile ad hoc networks. Two problems have to be solved: (1) how to learn about available data in the network; and (2) how to access desired data even when the original copy of the data is unreachable. In this paper, we develop an integrated data lookup and replication scheme to solve these problems. In our scheme, a group of mobile nodes collectively host a set of data to improve data accessibility for all members of the group. They exchange data availability information by broadcasting advertising (ad) messages to the group using an adaptive sending rate policy. The ad messages are used by other nodes to derive a local data lookup table, and to reduce data redundancy within a connected group. Our data replication scheme predicts group partitioning based on each node's current location and movement patterns, and replicates data to other partitions before partitioning occurs. Our simulations show that data availability information can quickly propagate throughout the network, and that the successful data access ratio of each node is significantly improved.

This paper propose a new multicast protocol for ad hoc wireless networks. The scheme, named Partial Flooding, increases packet delivery ratio at high mobility speeds and high traffic rate, without increasing the control overhead. The basic principle is that, nodes will switch to flooding when link failures happen between nodes that are involved in the multicast process. At high mobility speeds link failures are expected to increase and increasing the frequency of updates is not always the ideal way. Our scheme is adaptive to high and low mobility speeds.

In this paper, we present empirical study on new framework of location information service on the Web with CC/PP and RDF. We have implemented a prototype system and explored a new technique to exchange the location of a mobile device on the Web with CC/PP and its exchange protocol to localize a Web service. We have used CC/PP to describe user's context in the form of CC/PP and RDF, and exchange them via HTPP Extension framework to localize Web contents based on the terminal location. In addition, we have harmonized CC/PP with P3P to protect user's privacy and specify the information to achieve the services on a server side. We have investigated a new way for describing such kind of information with CC/PP and its protection mechanism by P3P. Finally, we have developed experimental systems to demonstrate effectiveness of our techniques. We believe CC/PP and RDF can be applied for describing user's environment, and P3P can complement the security aspect of CC/PP.

The generalization of the Internet and the recent technological developments in embedded systems and wireless networks contribute to the realization of a vision where access to information is possible at any moment and from anywhere. This is particularly attractive with information that is relevant to a specific geographic location. Applications that rely on location-based services to provide information to mobile users, or that support interaction with real devices in the user neighborhood, are called location- dependent applications and enhance the relationship between mobile users and a specific geographic location. However, the design of such applications breaks with the existing paradigms and methodologies as the mobile devices and the wireless communication infrastructures have characteristics that are very different from those of desktop computers and wired communication systems. The Around architecture is an open and extensible framework for location-based services that allows network services to be associated with specific geographic locations. By using this architecture applications can select services that are relevant for specific locations. Within the context of the Around project we have developed a prototype system with multiple location-based services and an application that accesses these services to provide information related to a town transportation system. This paper reports on the design and development of this location- based application. Its design raised several new issues, going from the computational model to the interface paradigm, which are also discussed in this paper. The developed application is composed of several modules: a set of agents which are autonomous units with the knowledge necessary to select and use location-based services in a specific thematic area (e.g. transportation); an HTML output area based on a browser metaphor; and a location-context module responsible for determining the user context. The results show that an application architecture based on a modular approach turns to be very flexible as it becomes very easy to extend its functionality by simply adding or changing the agents that deal with each location-based service.

In this paper, we describe the design and implementation of WebAlchemist, a prototype web transcoding system, which automatically converts a given HTML page into a sequence of equivalent HTML pages that can be properly displayed on a hand-held device. The Web/Alchemist system is based on a set of HTML transcoding heuristics managed by the Transcoding Manager (TM) module. In order to tackle difficult-to-transcode pages such as ones with large or complex table structures, we have developed several new transcoding heuristics that extract partial semantics from syntactic information such as the table width, font size and cascading style sheet. Subjective evaluation results using popular HTML pages (such as the CNN home page) show that WebAlchemist generates readable, structure-preserving transcoded pages, which can be properly displayed on hand-held devices.

A highly secure mobile agent system is very important for a mobile computing environment. The security issues in mobile agent system comprise protecting mobile hosts from malicious agents, protecting agents from other malicious agents, protecting hosts from other malicious hosts and protecting agents from malicious hosts. Using traditional security mechanisms the first three security problems can be solved. Apart from using trusted hardware, very few approaches exist to protect mobile code from malicious hosts. Some of the approaches to solve this problem are the use of trusted computing, computing with encrypted function, steganography, cryptographic traces, Seal Calculas, etc. This paper focuses on the simulation of some of these existing techniques in the designed mobile language. Some new approaches to solve malicious network problem and agent tampering problem are developed using public key encryption system and steganographic concepts. The approaches are based on encrypting and hiding the partial solutions of the mobile agents. The partial results are stored and the address of the storage is destroyed as the agent moves from one host to another host. This allows only the originator to make use of the partial results. Through these approaches some of the existing problems are solved.

Space-division optical switches are essential for the protection, optical cross-connects (OXCs), and optical add/drop multiplexers (OADMs) needed in future fiber-optic communication networks. For applications in these areas, we proposed a thermocapillarity switch called oil-latching interfacial-tension variation effect (OLIVE) switch. An OLIVE switch is a micro-mechanical optical switch fabricated on planar lightwave circuits (PLC) using micro-electro-mechanical systems (MEMS) technology. It consists of a crossing waveguide that has a groove at each crossing point and a pair of microheaters. The groove is partially filled with the refractive-index-matching liquid, and optical signals are switched according to the liquid's position in the groove, i.e., whether it is passing straight through the groove or reflecting at the sidewall of the groove. The liquid is driven by thermocapillarity and latched by capillarity. Using the total internal reflection to switch the optical path, the OLIVE switch exhibits excellent optical characteristics, such as high transparency (insertion loss: < 2 dB), high extinction ratio (> 50 dB), and low crosstalk (< -50 dB). Moreover, since this switch has a simple structure and bi-stability, it has wide variety of applications in wavelength division multiplexing (WDM) networks.

This paper outlines two ways in which liquid crystal devices can be used as optical switches in telecommunications networks. This has become a very important area of research as electronic switches are becoming bottleneck in both circuit switched and packet switched networks. These two applications have been highlighted as they require radically different switching techniques, yet liquid crystals can be used in both as the reconfigurable element. The first application is in circuit switching, where optical loss, crosstalk and polarization dependence are all critical parameters, whereas reconfigurration time is not. For this application a reconfigurable hologram written in phase with liquid crystal modulators is used to diffract optical beams in an optically transparent switch fabric. This technique does not suit packet switching as a much faster reconfigurration rate is required and loss and crosstalk are not critical. Hence a shutter based crossbar is used as a packet switch using a liquid crystal which can be reconfigured in 1 msec. A hybrid electronic/optical approach is also taken as this allows a much more scalable packet switch to be built.

The use of nonlinearly chirped fiber Bragg gratings (FBGs) for tunable dispersion compensation is described. It is experimentally demonstrated that wide dispersion tunability can be accomplished over multiple wavelength division multiplexed (WDM) channels simultaneously in a single device with an approximately 10 cm long nonlinearly chirped FBG. Multiple channels are generated in such a short grating by fabricating a periodic superstructure over (sampling) the FBG. Proper design of the sampling also allows one to achieve dispersion slope compensation over broad bandwidth in WDM communication systems.

This paper presents a comparative analysis between two specific post-processing techniques (RIE dry etching and TMAH wet etching) that are suitable for implementing a monolithic CMOS compatible MEMS fabrication technology. Further, an experimental investigation is presented which details the fabrication of MEMS structures by TMAH post etching of a CMOS chip fabricated in a standard AMI 1.5 micrometers CMOS process. Finally this paper provides future designers with experimental data that will allow for the design and fabrication of simple MEMS structures using a standard CMOS process.

We perform double sideband and single sideband suppressed- carrier modulators using a Mach-Zehnder interferometer in free space. Two bulk LiNbO₃ crystals are used to modulate the optical beam at 1319 nm in both branches of the interferometer. We present experimental results of the optical spectrum using heterodyne measurement technique. These results show maximum carrier and sideband suppression of 24 dB and 12 dB respectively for 1 GHz modulation frequency.

As IC densities grow to 100's of millions of devices per chip and beyond, the inter-chip link bandwidth becomes a critical performance-limiting bottleneck in many applications. Electronic packaging technology has not kept pace with the growth of IC I/O requirements. Recent advances in smart pixel technology, however, offer the potential to use 3-D optical interconnects to overcome the inter-chip I/O bottleneck by linking dense arrays of Vertical Cavity Surface Emitting Lasers (VCSELs) and photodetectors, which are directly integrated onto electronic IC circuits. Many switching and parallel computing applications demand multi-chip interconnection fabrics that achieve high-density global I/O across an array of chips. Such global interconnections require a high degree of space-variance in the interconnection fabric, in addition to high inter-chip throughput capacity. This paper reviews the architectural and optical design issues associated with global interconnections among arrays of chips. The emphasis is on progress made in the design and implementation of the second generation Free-space Accelerator for Switching Terabit Networks (FAST-Net) prototype. The FAST-Net prototype uses a macro-optical lens array and mirror to effect a global (fully connected) fabric across a 4 X 4 array of smart pixel chips. Clusters of VCSELs and photodetectors are imaged onto corresponding clusters on other chips, creating a high- density bi-directional data path between every pair of smart pixel chips on a multi-chip module. The combination of programmable intra-chip electronic routing and the fixed global inter-chip optical interconnection pattern of the FAST- Net architecture has been shown to provide a low latency, minimum complexity fabric, that can effect an arbitrary interconnection pattern across the chip array. Recent experimental results show that the narrow beam characteristics of VCSELs can be exploited in an efficient optical design for the FAST-Net optical interconnection module. A new design combines micro-, mini-, and macro-optical elements to achieve the required high registration and resolution accuracy while minimizing the packaging and alignment complexity.

We report on the fabrication of two-dimensional vertical- cavity surface-emitting laser (VCSEL) arrays designed for flip-chip integration with silicon complementary metal-oxide semiconductor (CMOS) circuits. Devices emitting at 980 nm wavelength show single-mode output powers of 3 mW, modulation current efficiencies up to 10 GHz/(root)mA, and are suitable for digital data transmission at 12.5 Gb/s. The chips are incorporated into optical area interconnect system demonstrators employing plastic optical fiber bundles as waveguiding medium. Substrate removal is applied as a possible route toward the production of modules compatible to the standard 850 nm wavelength regime. Fully pretested and mechanically robust arrays for short-wavelength bottom emission are alternatively demonstrated by providing light outcoupling holes in the GaAs substrate.

A challenging task facing the designers for the next generation of archival storage system is to provide storage capacities several orders of magnitude larger than existing systems while maintaining current data access times. To meet this challenge, a smart optoelectronic database filter suitable for large capacity relational database systems that use page-oriented optical storage devices has been developed. The photonic VLSI device technology based database filter monolithically integrates optical detectors, photoreceiver circuits, data manipulation logic, and filter control circuitry onto a single CMOS chip. Since only valid search data is passed to the host computer, the introduction of a database filter can dramatically reduce database search time. This paper presents the design and performance analysis of the database filter system. Simulation data suggested that a 32 X 32-bit filter fabricated in a 0.35 micrometers CMOS process could have an optical page read rate of 263 Mpages/s and support 286 Mrecords/s transfer rate to a host computer.

This paper presents a novel ATM router based on free space optical interconnects. The system consists of a VCSEL/CMOS chip that broadcasts N light beams (channels) which are divided into to M sections. A diffractive optic element then expands the number of these beams into N²/M beams which are incident on a smart photodetector array. This array performs an optical to electronic conversion and selects one of the 'input' channels and routes it to the proper output.

Indium gallium arsenide (InGaAs) photodiode arrays are used in a wide variety of optical communications-related applications. Two-dimensional arrays are used for laser beam profiling, assembly and performance monitoring of optical switches and add-drop multiplexers, and simultaneous aiming/detection for free space communications. Linear arrays integrated with self- scanned readout integrated circuits are used for the spectroscopic monitoring of WDM source arrays and for dynamic gain flattening of erbium-doped fiber amplifiers (EDFAs). Parallel output arrays are coupled with arrayed waveguide gratings (AWGs) both for power monitoring of WDM source arrays and direct detection of high-speed signals. In this paper we will summarize the status of InGaAs array technology and describe the various applications in detail.

The design, demonstration and evaluation of a general purpose, field programmable smart pixel based photonic information processing system is presented. This novel architecture incorporates programmable photoreceiver cells into a field programmable gate array (FPGA). Implemented with a photonic VLSI technology the device is suitable for prototyping photonic information processing systems. We report here on the programmable photoreceiver design methodology and measure device performance.

Residual arithmetic is a very recognized mathematical approach accepted in optical parallel processing for its inherent parallelism. Here in this paper we propose a new concept of using residue arithmetic in optical flip-flop or memory element.

We present a simple scheme for implementation of optical single sideband suppressed carrier (OSSB-SC) modulation based on the use of standard MZM and passive fiber optic components. The function of the proposed modulator is also theoretically predicted. A Sagnac interferometer structure is used for sideband suppression. The light propagating in one direction is orthogonally linearly polarized to the oppositely propagating light. The orthogonal polarization is implemented by a non-reciprocal optical element. The optical carrier is attenuated by appropriately setting the Mach-Zender Modulator (MZM) at quadrature bias point. We will experimentally demonstrate sideband suppression and carrier attenuation up to 28 dB by using all commercially available components.

In avionic systems, data integrity and high data rates are necessary for stable flight control. Unfortunately, conventional electronic control systems are susceptible to electromagnetic interference (EMI) that can reduce the clarity of flight control signals. Fly-by-Light systems that use optical signals to actuate the flight control surfaces of an aircraft have been suggested as a solution to the EMI problem in avionic systems. Fly-by-Light in avionic systems reduces electromagnetic interference hence improving the clarity of the control signals. A hybrid approach combining a silicon photoreceiver module with a SiC power transistor is proposed. The resulting device uses a 5 mW optical control signal to produce a 150 A current suitable for driving an electric motor.

The trend towards so-called digital convergence (multiple functionality within a single terminal) is opening up a need for high-capacity storage within the cellular mobile terminals (CMT). Solid-state memories and magnetic microdrives are the most commercially mature options. Optical disk technology in this size range is immature, but has a unique potential: no other medium at present has the capability to be simultaneously low-cost, high-capacity, and exchangeable. In this paper, we explore the requirements for the implementation of optical disk storage in a CMT environment. From the technical point of view, these requirements include small form factor, high-enough data density and throughput, low power consumption, robustness, low cost, mass productability, and modularity. Although current technologies may satisfy some of these requirements individually, there is a need for combined optimization of all of these parameters. From the commercial point of view, the most crucial requirement is global standardization. Such standardization is crucial if wide interoperability is wanted (between CMT manufacturers, and even more crucially between CMTs and other appliances). Current optical storage standards are industry-driven and tend to be proprietary and/or incompatible. Even if the technical challenges can be met, optical data storage is not likely to be accepted in CMT applications unless global standardization proceeds more quickly than it is doing at present.

Locomotive safety controlling system is a system currently used in railway locomotives. The system detects and acquires useful parameters, analyzes the data, gives right decision according to the results and saves information for future use. As three of the important functions, compression, storage and analysis of voice data have shown to b e extremely useful. The analysis process shows that joint time-frequency extraction is effective when restoring voice from the unclear data. Not only does the SNR of the reconstructed signal improve greatly, most components of some random noise severely blurred voice signal can be removed as well. Description of system scheme and data analysis is included in the paper.

For efficient business continuance and backup of mission- critical data an inter-site storage network is required. Where traditional telecommunications costs are prohibitive for all but the largest organizations, there is an opportunity for regional carries to deliver an innovative storage service. This session reveals how a combination of optical networking and protocol-aware SAN gateways can provide an extended storage networking platform with the lowest cost of ownership and the highest possible degree of reliability, security and availability. Companies of every size, with mainframe and open-systems environments, can afford to use this integrated service. Three mayor applications are explained; channel extension, Network Attached Storage (NAS), Storage Area Networks (SAN) and how optical networks address the specific requirements. One advantage of DWDM is the ability for protocols such as ESCON, Fibre Channel, ATM and Gigabit Ethernet, to be transported natively and simultaneously across a single fiber pair, and the ability to multiplex many individual fiber pairs over a single pair, thereby reducing fiber cost and recovering fiber pairs already in use. An optical storage network enables a new class of service providers, Storage Service Providers (SSP) aiming to deliver value to the enterprise by managing storage, backup, replication and restoration as an outsourced service.

Data access time is an important performance measure for any mass data storage system. It is the time between a data request received by the data storage system and the last bit of requested data sent to the host. For a magnetic hard disk drive, its data access time depends on how fast the actuator moves from tracks to tracks, the rotational speed of the platter, the size of the buffer for storing requests and data, and the algorithm that schedules the data requests. To improve the data access time, designers need to consider these parameters and perform trade-offs. This paper considers this problem. We derive analytical relations that are useful for design trade-offs.

In high-density magnetic recording systems, long channel memories and strong signal dependent noise are the two main factors that degrade system performance. The equalized Lorentzian channel can be seen to be equivalent to a partial response channel. As the density becomes higher, the length of the partial response becomes longer which makes the problem of achieving low bit error rates more challenging. The turbo principle is very promising for such channels, since equalization and coding can be combined together seamlessly. For simplicity, a serially concatenated turbo code is used here. A recursive systematic convolutional code is used as the outer code while the inner code is the modified partial response channel. In decoding, two a posteriori probability (APP) detectors are connected to form a loop. The outer code APP detector takes the output of the inner code APP detector as its input and then feeds back its output as the new input to inner code APP detector. This decoding procedure is carried on in an iterative manner. In decoding, it is often assumed that the noise component is white Gaussian for simplicity. We will consider signal dependent noise because it has the dominant effect when the recording density is high. We will demonstrate the improvement in bit error rate over the scheme that assumes white Gaussian noise.

The widespread deployment of wireless networks using the 802.11(b) standard across the country presents a rebirth of age-old network security problems along with a number of new ones. The wireless network, much like a shared network using broadcast devices such as network hubs, travels across a shared medium. Because of the structure any member of the wireless network can observe and intercept data being sent or received by other members. Unlike 'wired' networks there is no means to isolate traffic from other network members. The second security issue for wireless networks is the transmission of data 'clear text' so that if it is intercepted it can be read and used. Wireless networks bring about another problem that compounds the first two concerns that all shared networks must deal with, that is, anyone within the transmission range of the wireless network can join. No longer must a person enter a building to infiltrate a business network, they need only park across the street. The first implementation of network security for wireless was the WEP (Wired Equivalent Privacy) protocol. WEP attempts to make a wireless network at least as secure as a switched 'wired' network. The WEP protocol intends to secure the traffic integrity with the use of a RC4 cipher and a CSC-32 checksum. In the passphrase used for the RC4 encryption is also used as a form of access control. There are several critical faults in the WEP implementation that allow both passive data acquisition and active data modification. At 11 Mbit, capturing approximately 5 hours of clear text data can guarantee the capture of two packets with the same initialization vector (IV). Once the packets are used to get the clear text packet, that information can be used to decrypt any packets with the same IV. Since the IV's are only 24 bits the decryption of entire network becomes only an exercise in patience, with a 24 hours of continuous monitoring the WEP encryption can be defeated completely and a simple approximately 15 Gigabyte table of IV values can be used to encrypt any packet. On a network only partially infiltrated, bit-wise manipulated packets can spoof their validity using the linear nature of the CSC checksum. Any combination of passive and active attacks can be used to modify commands as they are being sent or login information can be taken for use on another network to access personal accounts. WEP also has a critical flaw outside of the sophisticated attacks that can be used to subvert its security. WEP uses a shared key known by both the client machines and the base stations. For this reason the key can be lost through human security problems. This includes the loss of equipment to theft, employee turnover and general mishandling of key information. In order to implement secure wireless networks it has become clear that a new scheme must be developed which can address the 3 security concerns mentioned earlier and at the same time function on existing hardware and software.