Presentation is loading. Please wait.

Presentation is loading. Please wait.

Three heuristics for transmission scheduling in sensor networks with multiple mobile sinks Damla Turgut and Lotzi Bölöni University of Central Florida.

Similar presentations


Presentation on theme: "Three heuristics for transmission scheduling in sensor networks with multiple mobile sinks Damla Turgut and Lotzi Bölöni University of Central Florida."— Presentation transcript:

1 Three heuristics for transmission scheduling in sensor networks with multiple mobile sinks Damla Turgut and Lotzi Bölöni University of Central Florida ATSN 2008 May 13, 2008

2 Introduction  Traditional sensor networks  Static, low-power, forward data by hop-by-hop routing, single or multiple sinks  Energy conservation  Alternative approach  Data collection by a set of mobile sinks  More economical for power consumption  Collect and buffer observations, transmit to them to the closest sink  Transmission scheduling problem: should I send the data now or wait for a more favorable moment?

3 Contributions  Describe and compare three practically implementable heuristic algorithms  H1: human-inspired simple heuristics  H2: stochastic transmission  H3: constant risk  Describe an optimal algorithm, based on a dynamic programming to provide a baseline for the comparisons  Not practical to implement

4 Transmission scheduling problem  Decision of the node whether to transmit or not its currently collected set of observations to mobile sink at a given point in time  Wait until mobile sink gets closer?  If wait too long, buffer may get full and loose data  If wait too little, may bypass better opportunities  Send it with lower power consumption?

5 Assumptions  Data transmission is initiated by the node  Mobile sink visits every node  All collected data may not be transmitted  Data transmission between the sensor node and the closest mobile sink  Sink does not move during transmission  No deadline with transmissions of data  Data buffering for an arbitrary amount of time without penalty

6 Objectives of the algorithms  Objectives of the nodes:  Transmit all the observations  Minimize the energy consumption  The scheduling strategy tries to minimize the objective function which balances these two factors  Energy minimization only, no observations may be transmitted  Data loss minimization only, transmission can occur at every opportunity

7 Cumulative policy penalty  Objective function: Cumulative Policy Penalty (CPP)  “Cumulative” aspect is essential here  Sum of the transmission energy + a penalty for lost packets  We can parametrize the relative weight of the lost packets  … but it can not be lower than the transmission energy… the node will improve its score by loosing all its packets!  Transmission energy is determined by the physical factors  The model used for energy dissipation used for communication

8 Related work  Routing towards mobile sink  SEAD (Kim et. al.), HLETDR (Baruah et. al.)  Mobility models of the sinks  Random, predictable, controlled  SENMA (Tong et. al.), Chakrabarti et. al.  Mobility and routing  mWSN (Chen et. al.), Luo et. al., Kansal et. al., Gandham et. al., Message ferrying (Zhao et. al.)  Transmission scheduling  Zhao et. al, Song et. al.  Combinations  Somasundara et. al., Guo et. al.

9 Oracle Optimal algorithm  Finds the optimal transmission schedule with the assumption that mobility patterns of the sinks is known  Optimality: find a schedule which minimizes the cumulative policy penalty for specified interval  Objective: serves the baseline for more realistic algorithms  Implementation: dynamic programming  Exponential in the worst case, in practice much faster

10 Three heuristics  Make their decision based on very simple calculations  Do not explore the solution space  Do not plan for the future transmissions  Notations used Mthe current buffer content M full the size of the buffer rdata collection rate d tr transmission rate dCurrent distance of the closest mobile sink

11 H1: Human-inspired simple heuristics  Mimic the human decision process for the transmission scheduling  Designed based on the observation of several humans  play the transmission scheduling problem as a game and then  describe their strategy  Humans are not comfortable doing calculations during the game

12 H1: Human-inspired simple heuristics (cont’d)  Strategies developed were based on levels of the buffer and the current distance of the mobile sink  Did not adhere strictly to the stated strategy  When asked, all agreed “coin toss” is not a good strategy

13 H1: Human-inspired simple heuristics (cont’d)  Parameters d opt Optimal distance MLML Too low to transmit MHMH Buffer emergency level  Algorithm

14 H2: Stochastic transmission  Transmits randomly with probability distribution affected by two factors  Level of buffer  Distance of the mobile sink  Final equation

15 H2: Stochastic transmission

16 H3: Constant risk  Estimate based on historical information how much risk a decision carry  Take decisions based on a constant risk factor  Goal: prevent the algorithm from being too bold in one occasion and too cautious in others  OP[t][d]: future probability

17 H3: Constant risk  Parameters p risk Constant risk factor tqtq Quantization of remaining time dqdq Quantization of the distance to the sink  Algorithm

18 Experimental Study  Performed a series of experiments using the YAES simulator framework  Scenario:  Mobile sinks are moving around collecting data from sensor nodes using one-hop communication  Random waypoint mobility pattern of the sinks

19 Simulation parameters

20 Compared implementations, measurements  Four different sensor implementations  Oracle Optimal (OrOpt)  Human inspired (H1: HI)  Stochastic (H2: STO)  Constant risk (H3: CR)  Measurements collected:  Total transmission energy  Data loss ratio  Cumulative policy penalty (CPP)

21 CPP w.r.t. transmission range

22 Consumed energy w.r.t. transmission range

23 Data loss ratio w.r.t transmission range

24 CMM vs. mobile sink count

25 Consumed energy w.r.t. mobile sink count

26 Data loss ratio w.r.t. mobile sink count

27  Investigated the problem of transmission scheduling  Agent approach where each node tries to maximize its utility by minimizing energy consumption and data loss  Presented an oracle optimal algorithm to provide a baseline for the comparisons  Described and compared three practically implementable heuristic algorithms  H1: human-inspired simple heuristics  H2: stochastic transmission  H3: constant risk Conclusions

28 Conclusions (cont’d)  Human intuition might lead us astray  Overall, the stochastic algorithm gave the best results, followed by constant risk  The human intuition inspired algorithm came out last  As expected, the oracle optimal algorithm provided the best results, but not by a wide margin

29 Thank you Questions?


Download ppt "Three heuristics for transmission scheduling in sensor networks with multiple mobile sinks Damla Turgut and Lotzi Bölöni University of Central Florida."

Similar presentations


Ads by Google