Introduction to blueweb: A decentralized scatternet formation algorithm for Bluetooth ad hoc networks

全文

(1)IEICE TRANS. COMMUN.,. VOL.E91-B,. NO.9 SEPTEMBER. 2008 2873. LETTER. IEICE/IEEE Joint Special Section on Autonomous Decentralized Systems Theories and Application Deployments. Introduction Algorithm. to Blueweb: for Bluetooth. A Decentralized. Formation. Ad Hoc Networks* Chih-Min. SUMMARY In this letter, a decentralizedscatternetformationalgorithm called Bluelayeris proposed. First, Bluelayeruses a designated root to constructa tree-shapedsubnetand propagatesan integervariable k1 called counterlimitas well as a constantk in its downstreamdirection to determinenewroots. Then each new root asks its upstreammasterto start a return connectionprocedureto convertthe tree-shapedsubnetinto a web-shapedsubnetfor its immediateupstreamroot.At the sametime,each newroot repeatsthe sameprocedureas the root to buildits own subnetuntil the whole scatternetis formed. Simulationresults showthat Bluelayer achievesgoodnetworkscalabilityandgeneratesan efficientscatternetconfigurationfor varioussizesof Bluetoothad hocnetwork. keywords: Bluetooth,ad hocnetwork,scatternetformation,networkscalability 1.. Scatternet. Introduction. Bluetooth is emerging as a potential technology for shortrange ad hoc wireless networks. This technology enables the design of low power, low cost, and short-range radio which is embedded in existing portable devices. Initially, Bluetooth technology is designed as a cable replacement solution among portable and fixed electronic devices. Today, people tend to use a number of mobile devices such as cellular phones, PDA's, digital cameras, laptop computers, and so on. Consequently, there exists a strong demand for connecting these devices into networks. As a result, Bluetooth becomes an ideal candidate for the construction of ad hoc personal area networks. Until now, many decentralized scatternet formation al-. YU•õa), Member. and. Chia-Chi. HUANG•õ•õ,. Nonmember. phases are designed to form a scatternet. In the first phase, multiple init nodes, each with the highest ID in its local neighborhood during the inquiry mode, are selected as root nodes (root selection process). Then each root pages its neighboring slaves to form piconets and builds its own subtree topology. In the second phase, the subtree topologies are merged into a single scatternet. Finally, Distributed Bluetree designates one of the subtree topologies as the root of the scatternet. To simplify the root selection process and make the scatternet topology controllable, a decentralized scatternet formation algorithm called Bluelayer is proposed in this letter. This method uses a designated root to propagate an integer variable k1, the counter limit, and a constant k in its downstream (outward) direction to determine new roots as well as build their associated subnets. The new roots are determined locally without exchanging information among nodes. With this method, the subnet size can be also controlled by appropriately selecting the constant k and each root manages its own subnet. This letter is organized as follows: In Sect. 2, we describe the scatternet formation algorithm of the Bluelayer. In Sect. 3, we use computer simulations to verify the system performance of Bluelayer. Finally, a conclusion is drawn in Sect. 4. 2.. Bluelayer Scatternet. Formation. Algorithm. gorithms for constructing a Bluetooth ad hoc network have been proposed [1]-[4]. In [1], a node which has complete knowledge of all the nodes is elected as the leader of the scatternet. Then, this leader partitions the entire scatternet topology by a predefined formula. In [2] and [3], multiple leaders each with the largest number of neighbors in its local neighborhood are elected as root nodes to form piconets. The leaders instruct their slaves to page specific neighbors in order to form gateways to neighboring piconets. In a Distributed Bluetree algorithm [4], two phases including the subtree formation and scatternet formation. At the beginning, a particular node is given as the designated root to set a counter limit k1=k, where k1 is an integer variable and k is the constant. With these two parameters, the first root inquires and pages up to 7 neighboring slaves, and forms its own piconet. Each slave then switches its role to master (called S/M node) to inquire and page one additional slave. After each S/M node connects to its slave, a role exchange mechanism is executed to make the S/M node function as a relay and make the slave function as a master. Then, these new masters decrease k1 by 1 and continue to. Manuscript received January 12, 2008. Manuscript revised April 18, 2008. The author is with the Department of Computer and Communication Engineering, Asia University, Taichung County, Taiwan. The author is with the Department of Communication Engineering, National Chiao Tung University, Hsinchu, Taiwan. *This work is supported by the National Science Council , Taiwan (NSC97-2218-E-468-006). a) E-mail: [email protected] DOI: 10.1093/ietcom/e91-b.9.2873. propagate the two parameters in the downstream direction. In this way, when the (k1)th master is reached, k1= 0. The master becomes a new root and the counter limit k1 is reset to k. The tree-shaped subnet of the designated root is created. Then this new root asks its upstream masters to start the return connection procedure and tries to connect with one additional slave until its immediate upstream root is reached. As a result, the tree-shaped subnet of the designated root is converted into the web-shaped subnet.. Copyright (c). 2008. The. Institute. of Electronics,. Information. and Communication. Engineers.

(2) IEICE TRANS. COMMUN., 2874. Fig. 1. The Bluelayer. At the same time, the new root repeats the same procedure as the designated root to build its own subnet and propagates the two parameters to determine new roots. This procedure is continued until the leaf nodes are reached. All the leaf nodes will request their immediate upstream masters to conduct the return connection procedure until its immediate upstream root is reached, and the whole scatternet is formed. Finally, each root manages its own web-shaped subnet. Here, we use k=2 in Fig. 1 as an example to describe the Bluelayer scatternet formation process. Initially, the designated root R1 inquires and pages slaves to form its piconets. Each slave then switches its role to master (called S/M node) to inquire and page one additional slave. After each S/M node connects to its slave, a role exchange mechanism is executed to make the S/M node function as a relay and make the slave function as a master. As a result, the R1 connects with the first tier masters, as shown in Fig. 1(a). There is a relay (slave/slave node) between R1 and its immediate downstream masters. The first tier masters decrease k1 by 1 and continue to connect with their downstream masters. When the second tier masters are reached and the counter limit k1=0, these masters become new roots and reset k1 to k, as shown in Fig. 1(b). The tree-shaped subnet of the designated root is created. These new roots ask their upstream masters to start the return connection procedure and connect with one addi-. scatternet. formation. VOL.E91-B , NO.9 SEPTEMBER. 2008. process.. tional slave until its immediate upstream root R1 is reached. The topology of the designated root is finished and it generates a web-shaped subnet. At the same time, these new roots start to page new slaves and connect with their immediate downstream masters (leafs in this example), as shown in Fig. 1(c), to build their own tree-shaped subnets. When the leaf masters are reached, these masters start the return connection procedure until their immediate upstream roots R2's are reached, and the scatternet formation process is terminated. Finally, all roots have their corresponding web-shaped subnet, as shown in Fig. 1(d). 3. 3.1. Bluelayer. System Performance. Simulation. Simulation Model and System Parameters. A simulation program is written to evaluate the system performance. First, we assume that the Bluetooth nodes are uniformly located on a rectangular lattice and the number of neighboring nodes which can be reached by each node is between 2 and 4. The simulated node number ranges from 60, 70, 80,, to 150. Two performance metrics, including the average number of roots and the average subnet size, are calculated by averaging over 100 randomly generated topologies for each simulated node number. The constant k is varied from 1, 2, 3, to 4 and the simulation results of Bluelayer are shown as follows..

(3) LETTER 2875. of is. roots. in. terms. number the. of. k=2. be. roots,. case. Figure Bluelayer. shows k=2. by and. the. inquiry. mode,. inquiry. packet.. In. calculated page,. return. on. Average. number. of roots. 3. control. master. mode,. the of. and. also. the. roots. can. selecting. an. overheads. packets. are. the. average. our. by inquiry,. simulation,. as. one. control to. the. calculated. inquiry. packets. perform. the. are. function. exchange.. that. than. both In. of. counted. cal-. for. function In. is. of. are. formation.. packets. packets. role. shows. overheads. overheads. connection.. the. number. page. Figure. by. control. execute. return of. the. addition,. of. size. scatternet. control to. page. by. There. as. number. total. average. during. and. the. average. In. formation. The. of. packets. cycle. .. the. performance. the. scatternet. average. response inquiry. size k,. size.. same. subnet. case.. mode. of. of. subnet the. various. the. the. value. Bluelayer.. number. page. number. average. a result,. for in. subnet. the. almost. value. 3. culated. each. the. As. with. inquiry. of. and. controlled k. average. among. Bluetree.. easily. largest. produces. appropriate. the. the. tradeoff. Distributed. Fig. 2. of. a performance. Bluelayer. spends. Distributed. much. Bluetree. in. less. electing. a. in a scatternet. root.. This. only. with. nodes. tion of. is. because. its. As. connected. a result,. overheads. determines. nodes. also. not. of. only. simplifies. formation. the. instead. Bluelayer but. scatternet. 4.. Bluelayer. to. the. root. node. neighboring. reduces. the. compared. all. root. root. selection. selecprocess. Distributed. Bluetree.. Conclusion. Bluelayer. is. a tier-based. spontaneously the. whole. ate. k. scatternet. value,. various. of. new. tributed. can. root. heads. in. value. appropriately.. cient. scatternet. selecting. is. it. As. an. configuration. for. than. Disover-. show. by. sizes. that. selecting. generates. various. pro-. control. results. Bluelayer. for the. efficient lower. scalability. a result,. appropri-. addition,. Simulation. network. and until. scalability. more. generates. roots.. better. roots. locally,. network In. process. new. achieves. By good. because. new. subnets. scatternet.. selection. electing. Bluelayer. generating own. formed.. achieve. Bluetooth. Bluetree. of their. is. we. size. posed. method. constructing. k. an. of. effi-. Bluetooth. scatternet.. References. Fig. 3. Average. control. packets. of scatternet. formation. [1]. T. Salonidis, topology Sel.. 3.2. Performance. Results. [2]. C.. P. Bhagwat, construction. Areas. Commun.,. Petrioli. and. formation. Figure 2 shows the performance on average numbers of roots for both Bluelayer and Distributed Bluetree. The average number of roots increases as the number of nodes increases and the average subnet size is defined as the number of nodes divided by the average number of roots. In Bluelayer, we observe that the number of roots decreases as k increases and the k=4 case produces the smallest number. for. pp. 222-226, [3]. [4]. C.. Comput.,. vol.52,. tional. Zaruba, to. Taiwan,. Nov. and. formation no.6,. I. for. pp. 779-790,. S. Basagni, enable. Conference. and. R.. personal. pp. 633-643,. networks,•h. S. Basagni, scatternet. G.V.. Tassiulas,. LaMaire, •gDistributed. area March. S. Basagni, •gDegree-constrained bluetooth. Multihop. mation. vol.23,. Taipei,. Petrioli,. L.. of Bluetooth. and. multihop. Proc.. IEEE. IEEE. Globecom. Communications,. J.. scatternet 2002,. vol.1,. 2002. Chlamtac, •gConfiguring Bluetooth June. networks,•h. BlueStars: IEEE. Trans.. 2003.. I. Chlamtac, •gBluetrees-scatternet. Bluetooth-based on. networks,•h 2005.. ad. hoc vol.1,. networks,•h pp. 273-277,. forIEEE. InternaJune. 2001..

(4)