Vishv Malhotra

An O(|V|3) algorithm for finding maximum flows in networks

An O(V cubed) algorithm is given for finding maximum flow in capacitated networks. THe algorithm is already well refenreces in literature and is being posted here to make it accessible to all in electronic form.

Optimisation in the Design of Environmental Sensor Networks with Robustness Consideration

This work proposes the design of Environmental Sensor Networks (ESN) through balancing robustness and redundancy. An Evolutionary Algorithm (EA) is employed to find the optimal placement of sensor nodes in the Region of Interest (RoI). Data quality issues are introduced to simulate their impact on the performance of the ESN. Spatial Regression Test (SRT) is also utilised to promote robustness in data quality of the designed ESN. The proposed method provides high network representativeness (fit for purpose) with minimum sensor redundancy (cost), and ensures robustness by enabling the network to continue to achieve its objectives when some sensors fail.

Resource and load sharing in mobile computing environments

Because of the mobility of the machines, it is difficult to provide a satisfactory solution to the problem of resource and load sharing. Mobile computers can change their position from time to time. In addition, their platforms are not only heterogeneous but also dynamic. This makes the resource and load sharing problem more difficult to tackle in such environments than in distributed systems. Finding a (near) optimal solution to the problem of load sharing in such environments is more complicated. Since the mobile computing environments are dynamic, traditional mechanisms for load sharing can not be employed without the support of new software or changes. This paper proposes a method for load sharing in mobile computing environments, using the abstract mobile-ticket engine in conjunction with migrations. Different migration mechanisms can then be employed in different situations to allow better level of load sharing.

An improved data-dependency-based backtracking scheme for Prolog

The conventional naive backtracking scheme for Prolog is inefficient. A more sophisticated backtracking scheme [l, 2, 3,4, 51 attempts to unprove the efficiency of execution by carefully selecting the backtrack points, without transgressing the depth-first backward-chaining, clause ordering or literal ordering within the clauses. However, many schemes [l, 3, 41, particularly those based on deductive analysis, are difficult to implement on a WAM (Warren’s Abstract Machine) [‘I]. &mar and Lin [5] have recently described a data-dependency-based backtracking scheme for Prolog that can be incorporated in a WAM implementation. We believe that the scheme, as presented, is not sufficiently selective; and as a result, it taps only a small part of the possible reduction in the search space far certain problems. For example, the 6-queens (clever) solution [l] shows only about 3.2% reduction in the search space. Assume a goal G fails to unify with a clausehead H. The search for a solution can now proceed in on,: of the following ways: Discard I-k Try to satisfy G by using a different clause. Discard G: Backtrack to an ancestor of G and try to search a solution without involving G.

Analysis of mobile authentication protocols by SVO logic

The next generation of mobile services makes it desirable for mobile users to be connected everywhere. Since these users usually in the move, roaming services are deployed to allow mobile users to access foreign network services without being limited to the geographical coverage of their home networks. Several solutions have been proposed based on either two-party or three-party roaming structure to allow ubiquities mobile access authentication, however, limitations such as performance issues and security vulnerabilities still exist in these approaches. In this paper, we implemented a methodology for verifying authentication protocols based on SVO logic, which shows that our Passport/Visa protocols meet the desired authentication objectives and prove the protocols correctness. Moreover, we show that our proposal addresses existing limitations when compared to the other approaches.

A Membership-Oriented Group Key Management for Application Services

Group key management is an important framework to secure group communications. To improve the performance of re-keying for group key management in large scale systems, logical hierarchical key was proposed. However, this approach does not consider the high re-keying cost of single user having multiple memberships. In this paper, we propose a membership oriented group key structure to optimize the re-keying of multiple membership users. Users in the systems are grouped into key-groups based on their memberships. Therefore, the re-keying operations for members of these groups can be combined without sacrificing forward secrecy. An analysis shows that the cost of re-keying operations in large scale multiple membership group key management can be improved with the proposed group key management.

An Algorithm to Use Feedback on Viewed Documents to Improve Web Query

This paper presents an algorithm to improve a web search query based on the feedback on the viewed documents. A user who is searching for information on the Web marks the retrieved (viewed) documents as relevant or irrelevant to further expose the information needs expressed in the original query. A new web search query matching this improved understanding of the users information needs is synthesized from these text documents. The methodology provides a way for creating web search query that matches the users information need even when the user may have difficulty in doing so directly due to lack of experience in the query design or lack of familiarity of the search domain. A user survey has shown that the algorithmically formed query has recall coverage and precision characteristics better than those achieved by the experienced human web searchers.

Storage-efficient data structure for large lookup dictionaries

We present a new structure, multiple bit hashing, for building dictionaries which is storage efficient and provides fast retrieval.

Do Managers Understand Importance of Securing IT Resources

Information infrastructures and resources has become critical component of the modern business and non-business organizations. In turn this dependence makes these organizations vulnerable to any significant failure in their information infrastructures and resources. Literature is full of examples of the companies suffering major losses and even demise as a result of information infrastructure and resources failures. To mitigate this vulnerability the senior management and governance of the organizations needs to pay direct role and attention to protect their critical information infrastructures and resources. This paper provides some results of a study we conducted recently to determine how the senior management of Malaysian business organizations view and control the information infrastructure and resources in their organizations to mitigate vulnerabilities to this critical component of their business organization.

Multi-objective optimization model in mobile data communication scheduling

The paper describes a multi-objective model to optimally schedule a service on a mobile device that requires data communication with the external data repositories.