Application-Level Middleware to Proactively Manage Handoff in Wireless Internet Multimedia

Paolo Bellavista, Antonio Corradi, Luca Foschini

Abstract

The proliferation of wireless multimedia services necessitates seamless connectivity during user mobility. Traditional IEEE 802.11 "hard handoffs" often result in significant latency and packet loss, disrupting real-time streaming. This paper presents MUM (Mobile agent-based Ubiquitous multimedia Middleware), a framework that utilizes mobile proxies and a proactive prediction algorithm to eliminate streaming interruptions. By preparing network paths and buffering data before a handoff occurs, MUM ensures continuous multimedia delivery across diverse Internet domains.

Introduction

In the contemporary Wireless Internet (WI) landscape, users demand high-quality multimedia streaming, such as Video-on-Demand (VoD), while roaming across heterogeneous networks. However, the IEEE 802.11 standard employs a "break-before-make" hard handoff approach. This process typically incurs a disconnection period of up to two seconds, leading to jitter, packet loss, and frequent re-buffering. While lower-layer solutions like Mobile IP address basic connectivity, they lack the application-level context required to manage multimedia-specific needs, such as dynamic quality adjustment based on available bandwidth or user subscriptions.

The MUM Architecture

MUM addresses these challenges through an application-level middleware approach. The architecture is built upon three primary components:

  1. Handoff Agents (HA): Mobile software proxies that track the user's location. When a handoff is predicted, the HA clones itself and migrates to a node near the target access point (AP).
  2. Proxy Switch (PS): An indirection point acting as an application-level router. It allows local redirection of data flows, shielding the media server from the client's mobility and reducing signaling latency.
  3. Two-Level Buffering: In addition to the client-side buffer, MUM implements a second-level buffer at the proxy. During the handoff interval, the proxy continues to accumulate data from the server, which is then "burst" to the client upon reconnection to rapidly refill the local buffer.
The MUM system architecture, illustrating the hierarchy of Domains, Subnets, and the placement of Proxy Switches (PS) and Handoff Agents (HA).

Proactive Handoff Management

The core of MUM's proactivity is the RSSI-GM Prediction algorithm. This lightweight mechanism uses Received Signal Strength Indication (RSSI) and a "Grey Model" to forecast imminent handoffs. This lead time allows the middleware to migrate agents and redirect traffic before the physical disconnection.

MUM categorizes handoffs into three types based on the network topology:

  • Micro-handoff: Movement between APs within the same subnet.
  • Macro-handoff: Movement between subnets within the same domain.
  • Global-handoff: Movement between different Internet domains.
Detailed signaling and data flow for (a) Micro, (b) Macro, and (c) Global handoffs within the MUM framework.

By operating at the application layer, the middleware can make intelligent adaptation decisions. For instance, if a user transitions from a high-bandwidth Wi-Fi cell to a lower-capacity network, the HA can proactively transcode the video or downscale it to audio-only to maintain session continuity.

Experimental Evaluation

To evaluate the MUM prototype, the authors developed a custom simulator capable of modeling complex roaming scenarios. The simulation environment included 2 Internet domains, 6 subnets, and 48 Wi-Fi cells, serving 600 simulated clients with varying movement speeds (0.6 m/s to 1.5 m/s). The traffic consisted of H.263-encoded VoD flows.

Key Results

The system's performance was measured across several metrics:

  • Prediction Efficiency: The RSSI-GM algorithm achieved up to 99% accuracy for global handoffs.
  • Streaming Continuity: Macro and Global handoffs maintained continuity at 99% and 95%, respectively, effectively eliminating visible interruptions for the user.
  • Setup Effectiveness: The system demonstrated 97-99% effectiveness in completing proxy setup and path redirection before the client reattached to a new AP.

While early predictions significantly improve continuity, they introduce a trade-off in storage overhead, as proxies must maintain buffers for longer durations.

Conclusion

MUM demonstrates that application-level middleware can effectively mitigate the limitations of hard handoffs in wireless multimedia. By decoupling endpoints through Proxy Switches and employing proactive agent migration, the system provides a seamless roaming experience. Future work will focus on extending MUM to support vertical handoffs (e.g., Wi-Fi to cellular) and implementing Service Level Agreements (SLAs) to manage traffic prioritization in congested environments.

Summary updated

Abstract

New deployment scenarios tend to consider the requirement of session continuity for service provisioning, especially multimedia streaming, to limited heterogeneous portable devices roaming among wireless localities. In particular, multimedia streaming should not experience any interruption while clients roam in wired-wireless integrated networks based on the standard best-effort Internet. The paper proposes an application-level middleware approach to proactively overcome Wi-Fi handoff and maintain multimedia session continuity in the wireless Internet by exploiting mobile proxies running on the wired network. Mobile middleware proxies locally support resource-limited clients, avoid packet losses during handoffs, pre-fetch local buffers with multimedia contents before handoff occurrence, and possibly reconfigure/renegotiate ongoing sessions after handoffs. Experimental results show that, notwithstanding the application-level implementation, mobile proxies can avoid streaming discontinuities with good efficiency in wireless-wired integrated networks even if their pro-activity is based on simple and lightweight handoff prediction techniques.
BibTeX
@inproceedings{bellavista2005application,
  author = {Paolo Bellavista and Antonio Corradi and Luca Foschini},
  title = {{Application-Level Middleware to Proactively Manage Handoff in Wireless Internet Multimedia}},
  booktitle = {MMNS},
  year = {2005},
  doi = {10.1007/11572831\_14},
}
Figure