Exploiting Social Interactions in Mobile Systems

This paper discusses a scheduling algorithm for the construction of bluetooth scatternets from existing piconets. A piconet is a collection of slaves controlled by a single master. The scatternet is the collation of multiple piconets.

Their proposed algorithm serves for the scheduling of both piconets and scatternets. It operates on the efficient simultaneous perturbation stochastic approximation gradient estimates (SPSA). In both cases they calculate the optimal threshold queue length values for allocating time slots to polled slaves. They then apply a heuristic for polling slaves that apparently provides significant performance gains.

I do not think this project has substantial level of relevance to our project. We are not experimenting with scatternets. Also the bluetooth networks that we will be working with in the context of gaming will be relatively small making a scatternet somewhat unnecessary.

http://dl.acm.org.lib-ezproxy.tamu.edu:2048/citation.cfm?id=1872874

An efficient algorithm for scheduling in bluetooth piconets and scatternets

This paper discusses a scheduling algorithm for the construction of bluetooth scatternets from existing piconets. A piconet is a collection of slaves controlled by a single master. The scatternet is the collation of multiple piconets.

Their proposed algorithm serves for the scheduling of both piconets and scatternets. It operates on the efficient simultaneous perturbation stochastic approximation gradient estimates (SPSA). In both cases they calculate the optimal threshold queue length values for allocating time slots to polled slaves. They then apply a heuristic for polling slaves that apparently provides significant performance gains.

I do not think this project has substantial level of relevance to our project. We are not experimenting with scatternets. Also the bluetooth networks that we will be working with in the context of gaming will be relatively small making a scatternet somewhat unnecessary.

http://dl.acm.org.lib-ezproxy.tamu.edu:2048/citation.cfm?id=1872874

WCA: A Weighted Clustering Algorithm for Mobile Ad Hoc Networks

The paper opens with a description of the demand and current state of mobile ad hoc networks. They then address their product, that is, a weighted clustering algorithm or WCA. They use this to restructure mobile adhoc networks when they determine the clusterheads, the most connected nodes, in a network. This algorithm determines the weight of paths between nodes for the purpose of increasing or lessening load levels. Its heuristics take into account a surprising number of variables including: ideal degree, transmission power, mobility, and battery power. It only triggers when it is needed, so as to not constantly restructure the network. It seems that in further distributing the load of the network, there comes a loss of connectivity.

This has a reasonably high degree of relevance to our project. Their algorithm takes into account many factors that are pertinent to mobile bluetooth gaming as players dropping and rejoining will necessitate a reorganization of their connections. Although in its current state, our system features a network where each node connects to every possible node.

http://www.springerlink.com.lib-ezproxy.tamu.edu:2048/content/knme635wbqmhvpx5/

WiFi and Bluetooth fight for bandwidth

This paper discusses the conflict that has arisen as a result of wifi and bluetooth operating on very similar bandwidths. They are close enough that they can often interfere with each other when used in closer proximity. While both experience interference as a result of this clash, bluetooth is far less succeptable. At most ranges between competing devices, bluetooth is relatively unaffected, but at ranges near 2cm severely dampens all bluetooth throughput. To get around this bluetooth has developed adaptive frequency hopping. Doing this they send test packets to detect congested frequencies and change paths to avoid collision. Wifi does not have this. But at close ranges it is a problem for both parties.

Techniques are being developed currently to allow bluetooth and wifi devices to exist in very close proximity. The most readily occurring of which was having the device, at driver level, control which was being broadcast. That way one radio device will be shut down while the other is broadcasting. This limits throughput however.

The paper did not contain a great deal of relevance to our project, as we are dealing with interactions of bluetooth devices at range, not optimizing hardware level issues. It was, however, interesting to note that bluetooth and wifi compete at all. It was a fact previously unknown to me.

http://www.edn.com/article/470440-WiFi_and_Bluetooth_fight_for_bandwidth.php

Social Networks and Mobile Games: The Use of Bluetooth for a Multiplayer Card Game

  This paper aimed to ascertain the social dimensions of multiplayer gaming. They conducted this experiment by designing the game BELKA. BELKA is a card game that was played both with physical cards in one test, and with HP Ipaq's in another. In the virtual gaming scenario, the users were allowed to move about the testing room.
  The interesting finding of this study was that a user's position relative to the bulk of the group can and did affect game play. They found that on average if a player was further from the table they made fewer moves. Basically the communal element of face to face gaming encourages the gaming activity itself.

http://www.springerlink.com.lib-ezproxy.tamu.edu:2048/content/q43d1ywux09rrn77/fulltext.pdf