Context Monitoring Framework for Sensor Nodes

Context Monitoring Framework for Sensor Nodes

1Avinash Palave,2 Prof. Dr.Shiv K Sahu,3Prof.Amit Sinhal

1Student, M.Tech.I.T. Technocrats Institute Of Technology Bhopal,M.P.India

2 Professor.HOD I.T., Technocrats Institute Of Technology Bhopal,M.P.India

3Professor,CSE Dept, Technocrats Institute Of Technology Bhopal,M.P.India

Abstract This paper can perform efficient processing and sensor control mechanisms by using context monitoring approach. Here implemented ESS algorithm for calculating ESS. Also implemented Greed MCFSS algorithm to reduce the energy cost as much as possible while simplifying the computation.This paper describes energy utilization by mobile sensor nodes.

By developing this paper can solve energy consumption problem. Here used performance metrics such as packet delivery ratio, end to end delay, energy consumption and additionally average throughput for comparison of seemon and existing system by using varying n/w scenarios and varying mobility.

Index Terms PDR , Essential Sensor Set, T-QSet, F- QSet, U- QSet,SeeMon,Packet Delivery Ratio

INTRODUCTION

In this paper, purpose SeeMon,a scalable and energy efficient context monitoring framework for personal context-aware applications. By using this context aware application, it continuously monitor users context and capture their changes over time.SeeMon

approaches the context monitoring problem by using two methods[1].

In first it uses unidirectional approach where the processing flow proceeds in one direction through a pipeline, which consists of several stages i.e preprocessing, feature extraction, context recognition and change detection.

In second it uses bidirectional approach, it forms a feedback path from the applications to sensors. This approach gives the opportunity to achieve a high degree of efficiency in computation and energy consumption.

1. SeeMon FRAMEWORK: – SeeMon is a middle-tier framework between personal context-aware applications and a personal sensor networks[1]. SeeMon provides programming APIs and a runtime environment for applications. , SeeMon receives and processes sensor data and controls the sensors in the personal sensor network. For the wireless communication between them, protocols such as Bluetooth and ZigBee can be used. In addition to wireless personal sensor network, device attached sensors such as accelerometer and gyroscope deployed on smart phones can be easily incorporated in the SeeMon framework without architectural change. SeeMon consists of four components: the CMQ

   Processor, the Sensor Manager, the Application Broker, and the Sensor Broker. Based on these components, the operation of SeeMon is performed in three phases: query registration, query processing, and sensor control.

2. SeeMon APIs:-The SeeMon has following APIs for proper functioning[1]

   registerCMQ(CMQ_statement) deregisterCMQ(CMQ _Id) createMAP(Parent_Map_Id) deleteMAP(Map_Id) InsertContextElement DeleteContextElement

3. CONTEXT MONITORING QUERY

   SeeMon provides CMQ, an intuitive monitoring query language that supports rich semantics for monitoring a wide range of contexts. It is important for applications to catch the changes in users context proactively.

   The CMQ template has the following format[1]: Context <context element>

   (AND <context element>)* ALARM <type> DURATION <duration>

   In the SeeMon calculating the ESS is a complicated problem. CMQs can be evaluated using a small number of sensors. It is simple to calculate ESS for the true- state CMQs and undecided-state CMQs. However, it is complicated to compute the set of essential sensors with minimum cost for the false-state CMQs. We call this problem minimum cost false-query covering sensor selection (MCFSS).By using this context monitoring query approach we can evaluate ESS Problem[i].

   Here also we have to use sensor control mechanism for controlling sensors based on the ESS calculation result. It sends a control message to the sensors that are not

   included in the calculated ESS. The control message configures the sensors to be put into the inactive state so that the sensors stop transmitting data. Afterward, the ESS Calculator updates the state of context elements related to the inactive sensors in the CMQ- Table.

   Table 1.CMQ Table

   Query ID	State	Context Element	Period	Time Stamp
   Q1	True	[f1.(bo,b1),true) [f2,(b4,b7),false)	12:00	11:45:12
   Q2	False	[f1.(bo,b1),false) [f2,(b4,b7),false)	Null	11:49:15
   Q3	True	[f1.(bo,b1),false) [f2,(b4,b7),true)	11:00	12:56:18

   CMQ Table can be represent as follows.

   In CMQ table we will get query id, state of query either it is true or false, period required for accessing query and time when query is executed as shown in table 1.

   This paper implementing ESS algorithm for calculating Essential Sensor Set. This paper will also employ Greed MCFSS algorithm to reduce the energy cost as much as possible while simplifying the computation in the broader sense. By using this above mentioned algorithm we can can easily solve ESS calculation problem and also reduce energy cost and time.

   Abbreviations and Acronyms

   If qi evaluates to false by sensor si, then RF- QCover belongs RF-QCover Union set of si.

   1. MCFQCover belongs Greedy-MCFSS

   2. ESS=TUQCover union MCFQCover.

      ABBREVIATION	ILLUSTRATION
      CMQ	Context Monitoring Query
      MCFSS	Minimum Cost False Query Covering Sensor election
      ESS	Essential Sensor Set
      T-QSet	Set of All True State CMQs.
      F-QSet	Set of All False State CMQs
      U- QSet	Set of All Undecided State CMQs.
      PDR	Packet Delivery Ratio

   3. Output is ESS

II.Greed MCFSS Algorithm.

F-QSet: a set of false-state CMQs.

S: Set of sensors,each of which covers a subset of F- QSet.

1. M = //Minimum cost of subset. S=S.

2. While F-QSet(M) subset FQSet do Find Sc

Where a(s) = COST(Si)

ALGORITHM

I.ESS Algorithm:-

S:a set of all true-state CMQs

T-QSet:a set of all true-state CMQs. F-QSet:a set of all false state CMQs.

U- QSet:a set of all undecided state CMQs.

q.sensor:a set of sensors which are associated with the context elements of a CMQ q

Steps:

1. TQCover,UQCover,TUQcover,RF-QCover=

2. For qi, where qi belongs T Q set. TQCover = TQCover Union qi.sensor.

3. For qi ,where qi belongs U -QSet. UQCover = UQCover Union qi.sensor.

4. TUQCover =TQCover Union UQCover.

5. RF-QSet =FQSet

6. For si ,where si belongs TUQCoverF

   If qi evaluates to false by sensor si,then RF-QSet belongs RF-QSet set of qi.

7. For qi where qi belongs RF-QSet

|F-QSet(Si) intersect F-QSet-F- QSet(M)|

S=S- set of Sc

RESULTS

In this paper basd on our aims and objectives, we are discussing the results from simulation studies. Here used different conditions to check the performance of proposed algorithm against the existing routing algorithms. Here used performance metrics such as packet delivery ratio, end to end delay, energy consumption and additionally average throughput. Following points explaining the every case and presenting their results:

1. Results According to Varying Network Scenarios In this first case, here used three network scenarios such as 10 nodes, 20 nodes, 30 nodes, 40 nodes and 50 nodes in order to evaluate the routing protocols performances. Network size is varying from 500×500 to 1000×1000 with constant simulation time 200-300 sec.

   1. Packet Delivery Ratio

      This is also one of the major performance metrics which evaluates the performance of routing protocols

      and TCP variants. packet delivery ratio is nothing but: PDR: total number of packets generated / total number of packets received. From the following graph 1A, the PDR of proposed robust routing protocol is better as compared to existing routing algorithms.

      GRAP 1A: PDR (%) Vs Data Scale

      GRAP 2A: Throughput Vs Data Scale

      400

      300

      200

      100

      Existing SeeMon

      120.00%

      100.00%

      80.00%

      60.00%

      Existing

      0

      10 20 30 40 50

      40.00% SeeMon

      20.00%

      0.00%

      10 20 30 40 50

   2. Average Throughput Performance

      Throughput is the ratio of total amount packets the receiver will receive from the source of the data within the specified time frame. End to end delay for the packet transmission is most important metrics for the throughput performance of the routing protocols. Following graphs showing the SeeMon framework for WSN having better throughput as compared to the existing context recognition based monitoring method.

      Table 2A. Average Throughput

   3. Total Energy Consumption

      In the simple case, the energy consumed by the network interface when a host sends, receives or discards a packet can be described using a linear equation

      Energy=m x size + b

      Table 3A. Total Energy Consumption Readings

      Nodes	Existing	SeeMon
      10	9597	9398
      20	10575	10058
      30	12398	11861
      40	17180	15863
      50	18923	17605

      20000

      15000

      Nodes	Existing	SeeMon
      10	143.95	171.92
      20	140.77	179.72
      30	247.25	315.62
      40	186.14	245.31
      50	204.57	230.66

      10000

      5000

      0

      10 20 30 40

      Existing SeeMon

      GRAP 3A: Energy (J) Vs Data Scale

   4. Average End to End Delay

      This metrics calculates the time between the packet origination time at the source and the packet reaching time at the destination. Here if any data packet is lost or dropped during the transmission, then it will not

      consider for the same. Sometimes delay occurs because of discovery of route, queuing, intermediate link failure, packet retransmissions etc are considered while calculating the delay. Such kind of metrics we have to measure against the different number of nodes, different traffic patterns and data connections.

      Table 4A.Average End to End Delay Readings

      1. Packet Delivery Ratio(PDR)

         GRAP 1B: PDR (%) Vs Mobility (M/S)

         105.00%

         100.00%

         95.00%

         90.00%

         85.00%

         Existing SeeMon

         GRAP 4A: Delay (Sec.) Vs Data Scale

         5 10 15 20

         Nodes	Existing	SeeMon
         10	0.957	0.8991
         20	0.905	0.895
         30	0.9759	0.75
         40	0.4887	0.4135
         50	0.6589	0.4756

         Mobility(M/S)	Existing	SeeMon
         5	97.98%	100%
         10	95.45%	99.83%
         15	90.83%	99.80%
         20	95.48%	99.92%

         Table 1B.PDR Reading

         1.5

         1

         Existing

      2. Average Throughput Performance

         0.5

         SeeMon

         0

         Table 2B.Average Throughput Performance

         Mobility(M/S)	Existing	SeeMon
         5	147.04	200.27
         10	143	198.66
         15	137.03	199.48
         20	143.63	199.25

         10 20 30 40 50

2. Results According to Varying Mobility/CMQ

In this second case performing the simulations based on varying number of mobility used. This used mobility varying in between 5 M/S to 20 M/S. Following the results presented for the same: Numbers of nodes are 10 with 500×500 network area and 200 seconds simulation time.

GRAP 2B: Average Throughput performance

better than existing system in respected to performance metrics.

1. Average Energy Consumption Table3B.Average Energy Consumption

   Mobility(M/S)	Existing	SeeMon
   5	72.49	33
   10	72.02	36.4
   15	72.41	40.7
   20	70.75	43.4

2. Total Energy Consumption

   Table 4B.Total Energy Consumption

   Mobility(M/S)	Existing	SeeMon
   5	724	331
   10	720	364
   15	724	407
   20	707	434

3. Average End to End Delay

Table 5B.Average end to end delay

Mobility(M/S)	Existing	SeeMon
5	0.9784	0.8993
10	0.953	0.8969
15	0.9069	0.8966
20	0.9534	0.8978

CONCLUSION

This paper implemented ESS and Greed MCFSS Algorithm.Here used performance metrics such as packet delivery ratio, end to end delay, energy consumption and additionally average throughput for comparison of seemon and existing system by using varying n/w scenarios and varying mobility.By using this paper can conclude that seemon framework is

REFERENCE DETAILS:

1. Seungwoo Kang,Jinwon Lee,Hyukjae Jang,Youngki Lee,Souneil Park,and Junehwa Song, A Scalable and Energy-Efficient Context Monitoring Framework for Mobile Personal Sensor Networks,IEEE Transaction On obile Computing,VOL 9,No.5,May 2010

2. T. Hofer et al., Context-Awareness on Mobile DevicesThe Hydrogen Approach, Proc. Hawaii Intl Conf. System Siences, 2003.

[3]. A. Rahmati and L. Zhong, Context-for-Wireless: Context-Sensitive Energy-Efficient Wireless Data Transfer, Proc. MobiSys, 2007.

1. S. Chakraborty et al., On the Effectiveness of Movement Prediction to Reduce Energy Consumption in Wireless Communication, IEEE Trans. Mobile Computing, vol. 5, no. 2, pp. 157-169, Feb. 2006.

2. G. Anastasi et al., Performance Measurements of Motes Sensor Networks, Proc. Intl Symp. Modeling, Analysis and Simulation of Wireless and Mobile Systems (MSWiM), 2004.

[6]. V. Shnayder et al., Simulating the Power Consumption of Large-Scale Sensor Network Applications, Proc. Intl Conf. Embedded Networked Sensor Systems (SenSys), 2004.

[7]. R.S. Sandhu and P. Samarati, Access Control: Principles and Practice, IEEE Comm. Magazine, 1994.

First Author

Mr.Avinash Palave completed B.E. (I.T.) From N.M.U, Jalgaon in 2006.Now presently doing M.Tech (I.T.) From Technocrats Institute Of Technology, Bhopal M.P.

Second Author

Prof.Dr.Shiv Kumar B.Tech(I.T.) From Bhagwat Institute Of Technology,Muzzaffarnagar U.P in 2006.He have received M.Tech(Honors,I.T.) from Technocrats Institute Of Technology, Bhopal M.P. in 2010.He have completed Ph.D.(C.S.E) from Banasthali University,Tonk(Rajashtan,India) in 2012.His research area interest includes voice signal compression, Image Processing, Voice Recognition. He is having 30 papers/articles published in International Journals.

Now currently He is working as Professor and HOD.IT in Technocrats Institute of Technology, Bhopal since November 2006.

Third Author

Prof. Amit Sinhal completed his B.E. in Computer Engineering from NIT Surat in 1996, M.Tech in Computer Science & Engineering from SATI Vidisha in 2005 and pursuing Ph.D. from Rajiv Gandhi Technical University, Bhopal. He worked in various reputed software development companies as Project Lead and University Institute of Technology, Barkatullah University Bhopal as Assistant professor.

He is having 15 papers/articles published in International Journals. Currently he is working as professor in Technocrats Institute of Technology, Bhopal in CSE Department