Essay: Implementing Swifinet

ABSTRACT
Use of wireless sensor networks (WSN) has resulted in many revolutionary changes in human life. WSN has gain significant concentration from scientist and end users. ‘SWiFiNet’ is task distributed reusable system architecture. In this architecture complex functionality has been transfer to the second tire devices of the system. Second tire devices are provided with more resources. ‘SWiFiNet’ satisfy the desire architecture of the WSN. This paper aims to analyze various reusable wireless sensor networks and concept related to ‘SWiFiNet’. And it provides the existing architecture of ‘SWiFiNet’ and also the functionality of the component of the reusable network architecture. This paper also provides idea about the proposed ‘SWiFiNet’ which will be developed on IEEE 802.15.4 MAC/PHY layer which will be having more advantages over its counterparts.
Keywords:- WSN, SWiFiNet, Reusable architecture.
I. INTRODUCTION:
Use of wireless network is increasing in every field, day by day. WSN is extensively use where wired network cannot be deployed or is expensive to deploy. WSN opens many ways for research community to further enhance wireless communication efficiency and productivity. Many protocols and algorithms has been developed to addressed many kind of issues related to WSN. But most of these protocols and algorithms concentrate on routing, energy efficiency, reliability and congestion control.
No more research is done on creating some integrated network architecture that will make the implementation of any WSN application easy. It means to make the components of the WSN system reusable. The main focus of WSN is on creating more and more efficient wireless interfaces. The wireless sensor network generally developed for specific work. These work include climate reporting, military applications, fuel level indicator and many more. While creating such system one has to start from the scratch and it becomes burdensome.
So, for all these difficulties, one solution is to create reusable WSN framework. This kind of framework is developed in ‘SWiFiNet’. It is a task distributed System Architecture for WSN.
This paper provides overview of ‘SWiFiNet’ and reusable architecture for WSN. In first section provided introduction about WSN and ‘SWiFiNet’. Second section provides related work regarding task distributed network architecture. Third section provides design goals of reusable architecture. Fourth section provides WSN architecture along with ‘SWiFiNet’ architecture comparison. Fifth and last part concludes the paper.
II. RELATED WORK
There has been attempts to make hierarchical task distributed reusable wireless sensor network architecture. Most of such proposals were developed using 802.11. The different applications related to reusable WSN architecture is as follows:-
ART wise gateway architecture was presented by Leel et al [4]. The architecture is two tire architecture. Tier one uses IEEE 802.15.4 and tire two uses 802.11. Linfeng develop Environmental adaptive Architecture model for wireless sensor network [5]. It was two tire architecture. The second tire was capable of adding number of nodes without affecting two tire architecture. ANGLE [5] architecture was develop for the health care applications. In this application authors claims that it can be reused for any other application.
Open and reconfigurable wireless sensor network has been proposed by Triantafyllidis. It was developed for pervasive health monitoring. Its main emphasis was on easy extension with additional sensor functionality. ZUMA [6] was developed for centralize smart phone platform which will interconnects all kinds of smart phone devices.
The ReWINS [3] research initiative is an attempt to advance such an end-to-end solution with support for incremental arrangement through a transparent lower layer implementation and control architecture and a user-friendly application interface.
Wireless Integrated Network Sensors or the WINS [7] project and NIMS project at University of California, Los Angeles is about ad-hoc wireless sensor network research dealing mainly with constructing micro-electronic mechanical sensors (MEMS) [8], effective circuit design, and design of self-organizing wireless network architecture. Though these projects have been successful in demonstrating a network of self-organized sensor wireless nodes, they seem to have a bias towards environmental and military applications. Also they use proprietary RF communication technology and hence the solutions are limiting for wide scale deployments in industries.
Motes and Smart Dust project [9] at University of California, Berkeley involved creating particularly low-cost micro-sensors, which can be suspended in air, buoyed by currents. Crossbow Inc. has commercialized the conclusion of this project. Here again the solution is limiting, as exclusive communication technologies have been used to achieve inter-device communication. Further, the focus has been on development of sensors and their communication rather than how the sensors will be integrated to form systems. This is generally termed as the ‘bottom-up’ approach, which may not be suitable for building complex systems.
Pico-Radio [10] ‘ A group headed by Jan Rabaey at University of California, Berkeley is trying to build an integrated wireless application interface called Sensor Network Service Platform. An attempt is to develop an interface that will abstract the sensor network and make it transparent to the application layer. A introductory draft describing the application interface has been recently released. They believe in a ‘top down approach’ (from control to sensor nodes) for building sensor networks which is probably more suitable for building complex systems.
Recently, there have been several work initiatives like TinyDB [11], Cornell’s Cougar etc. to develop a declarative SQL-like language to query sensors and define certain standard query services. Here the employment is sensor-interface specific and not a general or abstracted sensor networking platform. These query services can be implemented with ease on top of our (developed) wireless interface and sensor networking platform and can be made generic by extending them for other sensors.
Other research initiatives in this field include MIT’s ??AMPS, Columbia University’s INSIGNIA, Rice University’s Monarch. Though there have been a lot of research efforts in developing ad-hoc wireless networks, the focus has been on developing smart wireless sensor interfaces and not much attention has been paid to the actual application integration. Typical approach has been to develop powerful smart wireless interfaces, which supports the important features/requirements for a particular class of applications (like military, environment sensing or more focused applications like fuel-level control in automobiles). The result is a number of wireless interfaces appropriate for a certain class of application; but almost no interoperability between them. We believe that the deployment of wireless infrastructure in industries will occur in incremental stages and thus interoperability (between different sensor-networks) and extendibility (according to application needs) will form the basic requirements of any prospective solution. A prospective good solution would be an end-to-end solution, which is modular and extendable.
All of the above applications were the forward steps towards the making of reusable network architecture. From the comparison of the above all network architecture, it is clear that user of hierarchical network give advantage over flat network. And three tire architecture is also better for reusable WSN architecture.
EXISTING SYSTEM:-
‘SWiFiNet’ [1] is task distributed and reusable component architecture. The task layer hierarchical model of ‘SWiFiNet’ is as shown in the given figure. It follows hierarchical architecture comprising of Master node, router node and sensor node. The base layer is 802.15.4 MAC/PHY layer.
Table 1. Comparison between different reusable wireless sensor network applications
Sr. No Application/Project Purpose of Application/Project Limitations
1 ART gateway architecture To develop Hierarchical two tire architecture each tire having different base MAC/PHY The system is more complex due to different protocol in each layer.
2 EAWNA To develop tire two in such a way that adding more sensor dose no affect the architecture The focus is given more on second tire of architecture.
3 ANGEL For health care application. Claimed to be reusable Reusable for specific health care application
4 ReWins To develop such an end-to-end solution The main focus is given on industrial scenarios.
5 Wireless Integrated Network Sensor Building efficient circuit design, and design of self-organizing wireless network architecture These project seems to have bias towards environmental and military application
6 Motes and smart dust project Creating low cost micro sensor which can suspended in air The focus is on sensors rather how they will be integrated
7 Pico-Radio To make sensor network transparent to application layer Suitable for building complex systems only
8 TinyDB To develop declarative SQL-like language to query sensor Not a generic or abstracted sensor networking platform
.
The upper layer to this layer is DLL task layer. All the network communication is controlled by this layer. The functionality of different device component of the architecture is as given below:
Sensor node: The task of sensor node is restricted to sensing background and connecting the neighboring nodes or the master node. When the packet is handed over to any parent node, then it is the responsibility of the parent node to transfer the packet to the master node. In this node’s layer architecture, the network layer will have the addresses of the parent devices. It will be stored in the table. Two types of addresses are present in the table. First one is primary parent address and the other is secondary parent address. If primary address parent is not available then secondary parent address will be try by the node.
If network layer don’t contain any table then a ‘Hello’ packet request is broadcast into the network. Then the neighboring device will issue the join request. The join request will be accepted if the device meets LQI threshold criteria.
Router Node: – Router nodes are come in use when sensor nodes are not in the range of Master Node. Clustering is also possible using router node. In ‘SWiFiNet’ distributed architecture the functionality of the router node increased remarkably. The complex implementation of any application or protocol will be deployed on the second tire devices. The router node network layer also maintains the table containing entries of the parent devices to the router node. It can be router node again or router will send the information directly to the master node. Router nodes will generate join request if they receive hello packet request from any sensor node.
Master node: The complete topology information of the network is maintain by the master node. When there is queried transmission model, master node will generate query and send it in the network. Whenever the routing tables are updated in end nodes or router nodes will send the information to the master node.
Sensor node Router Node Master Node
Fig 1. Distributed architecture of ‘SWiFiNet’
III. PROPOSED MODEL:
The ‘SWiFiNet’ model developed [1] is successfully implemented on the hardware as the architecture of the ‘SWiFiNet’ is defined. But ns-2 simulation of ‘SWiFiNet’ is based on 802.11a. As the basic idea of MAC/PHY layer is to have 802.15.4 which will work more efficiently on the ‘SWiFiNet’ architecture. The architectural diagram [1] is as shown below:-
In many of the sensor network architecture the sensor nodes are based on the 802.11a, but in ‘SWiFiNet’ MAC/PHY layer is based on 802.15.4. It is having low bandwidth comparative to the 802.11a. As the ns-2 modulation of the architecture is done using 802.11a, we cannot compare the exact result of hardware simulation and the software simulation.
Dynamic source routing (DSR) has similar characteristics as of ‘SWiFiNet’. After comparing the hardware results [1] with DSR it is found that ‘SWiFiNet’ is having better performance against many parameters. Implementation of ‘SWiFiNet’ using ns-2 on 802.15.4 IEEE is not yet checked against other protocols and hardware results.
The software architecture [2] of the ‘SWiFiNet’ agent is as shown in figure 2. This diagram shows the network component simulation in ns-2. The devices can be identified by the variable passed from TCL script. 1 value represent master node. 2 value represent router node and 3 value represent sensor node.
Fig 2: Component simulation in ns-2
IV. Conclusion:-
The survey paper provides a detailed comparison and description of the well-known projects and applications which are developed from reusable architecture point of view. ‘SWiFiNet’ is basically a task distributed generic reusable architecture for WSN. The functionality of sensor node is reduced and has been transferred to the second tire devices like router node. By doing so sensor nodes are restricted to gather information and connecting to neighboring nodes. As time synchronization overhead is not present the lifetime of the network increase.
This reusable architecture can be used for various range of applications and can be configured accordingly. This paper gives proposed ‘SWiFiNet’ system which will be based on 802.15.4 IEEE standard and will be useful for researchers in many ways.
V. REFERENCES:
[1] A. H. Willig, J. H. Karowski, N. Baldus, H. Huebner, A., “The ANGEL WSN Architecture,” in Electronics, Circuits and Systems, 2007. ICECS 2007. 14th IEEE International Conference on, 2007, pp. 633-636.
[2] A. W. Rohankar, Mrinal K. Naskar, Amitava Mukharjee, ‘SWiFiNet’: Task Distributed System Architecture for WSN’ in IJACSA Special Issue on Selected Papers from International Conference & Workshop On Advance Computing 2013.
[3] A. W. Rohankar, Mrinal K. Naskar, Amitava Mukharjee, ‘A step towards reusable WSN architecture’, International Journal of Research and Reviews in Wireless Sensor Networks (IJRRWSN) vol. Vol. 2, 2012.
[4] B. S. P. Harish Ramamurthy, Rajit Gadh, “Reconfigurable Wireless Interface for Networking Sensors (ReWINS),” in proceeding of the 9th IFIP International Conference on Personal Wireless Communications (PWC 2004), 2004.
[5] J. C. Leal, A. Alves, M. Koubaa, A., “On a IEEE 802.15.4/ZigBee to IEEE 802.11 gateway for the ART-WiSe architecture,” in Emerging Technologies and Factory Automation, 2007. ETFA. IEEE Conference on, 2007, pp. 1388-1391.
[6] L. Linfeng, “Research on Environment-Adaptive Architecture Model of Wireless Sensor Networks,” in Networks Security Wireless Communications and Trusted Computing (NSWCTC), 2010 Second International Conference on, 2010, pp. 130-133.
[7] M. N. K. V. G. Soini, J. Rabaey, J. M. Sydanheimo, L. T., “Beyond Sensor Networks: ZUMA Middleware,” in Wireless Communications and Networking Conference, 2007.WCNC 2007. IEEE, 2007, pp. 4318-4323.
[8] Micro-Adaptive Multi-domain Power-aware Sensors (??AMPS) project at University of California, Berkeley, URL: http://www-mtl.mit.edu/research/icsystems/uamps/
[9] Pico-Radio project at University of California, Berkeley,URL:http://bwrc.eecs.berkeley.edu/Research/Pico_Radio/
[10] Smart Dust and motes project at University of California, Berkeley, URL: http://robotics.eecs.berkeley.edu/~pister/SmartDust/
[11] ] TinyDB project at University of California, Berkeley, URL: http://telegraph.cs.berkeley.edu/tinydb