7/31/2019 Oppnet Poster INSS06

http://slidepdf.com/reader/full/oppnet-poster-inss06 1/1

INSS 2006 1

 Index Terms—Opportunistic network, oppnet, sensor network, res-

cue and recovery, security, privacy. 

We present a new paradigm of  Opportunistic Networks, or

Oppnets. To the best of our knowledge, we have been the first

to define and investigate oppnets [1, 3]. Oppnets will enable

new or improve known applications—most notably in the area

of emergency preparedness and response [2].

Oppnets differ from traditional networks, in which the nodes

are all deployed together, with the size of the network and

locations of  all its nodes pre-designed (at least the initial

locations for mobile networks).

In oppnets, we first deploy a seed oppnet , which may be

viewed as a pretty typical ad hoc network. It self-configures

itself, and then works to detect “foreign” devices or systems

using all kinds of communication media—including Bluetooth,

wired Internet, WiFi, ham radio, RFID, satellite, etc. At this

stage oppnets starts to differ from typical networks. (Figure 1

depicts basic operational phases in an oppnet.)

Detected systems are identified and evaluated for their

usefulness and dependability as candidate helpers for joining

the oppnet. Best candidates are invited into the expanded 

oppnet . A candidate can accept or reject the invitation (but in

life-or-death situation it might be ordered to join). Upon

accepting the invitation, a helper is admitted into the oppnet.The resources of the admitted helper are integrated with the

oppnet, and tasks can be offloaded to or distributed amongst

this and all other helpers. A decentralized command center—

either augmenting human operators or fully autonomous—

presides over the operations of the oppnet throughout its life.

As an illustration, consider the following scenario. After a

visual detection of an overturned car by a security surveillance

helper, the second helper uses image recognition to read its

license plate, the third helper connects with OnStar system,

and via it contacts with two candidate helpers: Body Area

Networks (BANs) on (or within) bodies of two car occupants.

The BAN helpers tell the oppnet whether help for theirwearers is needed.

After the goals of the oppnet have been realized, each helper

must be released and restored to the state that is the closest to

the state that the oppnet admitted it in, thus minimizing

intrusiveness of helper’s participation in oppnet’s tasks.

Helpers provide access to communication, sensing,

computational, storage and other capabilities that would not be

available otherwise. In particular, oppnets can be used as a

bridge between disjoint communication systems, and a way to

leverage a huge set of “foreign” resources and services.

Our goal is to utilize results from related research and adapt

them for oppnets. Growth in oppnets most closely resembles

growth in unstructured peer-to-peer (P2P) systems. For

detection of helpers, oppnets might mimic ways used by

worms and viruses to spread. Grid computing can contribute

techniques for integration and management of heterogeneous

nodes and devices.

Fig. 1. Basic operational phases in an oppnet.

We present the numerous research challenges present in

oppnets, and initial solutions to some of them.

REFERENCES 

[1]  B. Bhargava, L. Lilien, A. Rosenthal and M. Winslett, “Pervasive

Trust,” IEEE Intelligent Systems, vol. 19(5), Sep./Oct.2004, pp. 74-77.

[2]  L. Lilien and A. Gupta, “Opportunistic Networks for Emergency

Preparedness and Response” (submitted for publication).

[3]  L. Lilien, Z.H. Kamal, V. Bhuse and A. Gupta, “Opportunistic

Networks: The Concept and Research Challenges,”  Intl. Workshop on

 Research Challenges in Security and Privacy for Mobile and Wireless

 Networks (WSPWN 2006) (to appear).

Opportunistic Networks

Leszek Lilien1, Zille H. Kamal

3, Ajay Gupta

1, Vijay Bhuse

3, and Zijiang Yang

2 

1Senior Members,

2 Member and 

3Student Members, IEEE 

WiSe (Wireless Sensornet) Lab, Department of Computer Science

Western Michigan University, Kalamazoo, MI 49008-5466 

YES 

NO 

Collaborative

processing 

Detect candidate

helpers (system)

GROWTH 

Integrate helper’s

resources

Admit candidates

into oppnet

Restore and release all

helpersNOYES

Decentralized

Command

Center

Need to

grow?

YES

Deploy

seed oppnet

NO

Enough

good

candidates 

Work

done?