Abstract

Users of handheld wireless devices do not have the luxury to wade through pages and pages of information to find the products and services they need. In addition, the inconveniences of using the thumb controls on a cellular phone to navigate, along with the high costs of wireless data network access highlights the problems that wireless users face. The need for direct access to highly relevant, location-aware information with a minimal amount of browsing and a minimal amount of user input is what prompted us to investigate the feasibility of and development of a plan for deploying mobile agents to perform information filtering and retrieval on behalf of the users in a mobile commerce setting. In this paper, we discuss a novel mobile agent platform that can be used for comparison shopping in a mobile wireless environment, through which businesses can better understand and communicate with the mobile consumer. The proposed agent-based approach helps the end user reduce information overload by helping her filter out unnecessary information and only show information that is relevant; it provides novel ways for cellular service providers to compete, and businesses to acquire new customers. We will address the challenges associated with such an evolutionary approach, and provide some sample scenarios, along with an architectural design for the proposed system. A proof-of-concept prototype using Java RMI has been constructed as part of the Havana mobile agent platform.

Keywords: Havana; Mobile agents; Software agents; Closed architecture; Business model for agents; Mobile commerce; Location-aware advertisements

Article Outline

1. Introduction

2. Challenges in mobile commerce

3. Mobile agents

4. The Havana agent platform

4.1. Business model

5. Comparison shopping agent

5.1. The advertisements module

5.2. The user location module

5.3. Life-cycle

6. Implementation and experiments

6.1. Sample operating scenario

7. Challenges and possible solutions

7.1. Acquisition of the location of the wireless devices

7.2. Protecting the user’s privacy

7.3. Ease of deployment

7.4. Mobile agent security

7.5. User resistance

8. Related work

9. Conclusion and future work

Acknowledgements

References

Vitae