Intelligent Content Transformation in Tatical Wireless Networks

Caruso et al

content transformation filtering manet tactical edge c2mincs

@article{caruso:milcom-2010,
  title={Intelligent Content Transformation in Tactical Wireless Networks},
  author={Caruso, F. and Chen, T. and Fiuk, M. and Kaul, V. and
          Hokelek, I. and Parmeswaran, K. and Sultan, F. and
          Samtani, S. and Shur, D. and Bowcock, J. and Thomas, S.},
  booktitle={Military Communications Conference ({MILCOM})},
  year={2010},
  pages={887--892}
}

MANET nodes have widely variable patterns of bandwidth available

Heterogeneous Intelligent Filtering (HIF) brokers autonomously transform data to match conditions across multi-domain networks

• XMPP chat, SOA Web Services, VOIP

Driven by QOS feedback from the network

Most effective at the gateways, where discontinuities in capabilities largely exist

Network-based data transformation

Most real time services adjust to user user demand and available resources through admission control

• Not applicable in some circumstances
• HIF applies data reductions instead

Conducts active probing to estimate bandwidth and loss

Filtering needs to happen with some cognizance of application semantics

• Must account for application and domain attributes such as location, roles, priorities

Transcodes video using VLC

Persistence module supports replay of several seconds of video

Selectively drops c2mincs traffic based on filtering rules such as raidus

Adapts XMPP to run over Reliable UDP

• TCP congestion control acts adversely under typical MANET loss rates
• Blocks packets based on keywords in content, markers