200 likes | 281 Views
Scheduling Messages with Deadlines in Multi-hop Real-time Sensor Networks. Wei Song. Introduction.
E N D
Scheduling Messages with Deadlines in Multi-hop Real-time Sensor Networks Wei Song
Introduction • “A cyber-physical system (CPS) integrates computing, communication and storage capabilities with monitoring and / or control of entities in the physical world, and must do so dependably, safely, securely, efficiently and in real-time.” - S. Shankar Sastry, UC Berkeley • RTSS 08 specialized track: Cyber-Physical Systems • RTSS 07 Architecture and Composibility for Cyber-Physical Systems • NSF Workshop on Cyber-Physical Systems was held on October in Austin, Texas 2006
Two features make it different from Today’s Sensor Network • control of entities • real-time Y. Tan , SIGBED Review, 5(1), 2008
Application • RTSS 08 Student Design Competition http://www.ieeta.pt/lse/ciberRTSS08/ a simulated multi-robot rescue mission (simulated CPS system)
RTSS 08 Robot Competition (Simulated CPS System) • Sensor sub system is ready to use • Actuator sub system is ready to use • Strategy Computation and simulated wireless communication need to design • Target location algorithm • Shortest path algorithm and path planning • Real time sensor request scheduling • Real time wireless communication (what? How?)
Scheduling Messages with Deadlines in Multi-hop Real-time Sensor Networks* • Application Background • Definitions and Notations • Algorithms • Example • Conclusion *Huan Li et al. “Proceedings of the 11th IEEE Real Time and Embedded Technology and Applications Symposium” (RTAS 05) 1080-1812
Application Background • Consider a team of robots that collaborate to achieve a common task. • Examples: searching for trapped people in a building on fire; building a map of an unknown environment • The robots are equipped with sub set of numerous sensors such as camera, temperature, pressure, infrared sensors, and positioning devices. • Each robot is also equipped with a wireless connectivity, communicate with others over an ad-hoc network.
Different Roles and Communication Requirements • each robot in the team may be assigned one or more tasks • leader-follower groups • the leader robot is designed with a task of determining a “plan” • a communication plan who is talking to whom • a path plan where to move
Definitions and Notations • Data Validity • Time interval for which data value produced by sensor is valid • sensor period • sensors produce data periodically • Effective Deadline (ed) • a sensor value is produced at time t, and its validity is v time unit. δis start time of the consuming task that consumes m • Latest Start time (LST) • latest time by which a hop must start transmitting a message for it to reach its destination by effective deadline
Definitions and Notations • mi denote the transmission of message m at the i hop • pd(mi) denote the total propagation delay and transmission time that will be incurred on the remaining hops to the destination • LST(mi) = ed(m) – pd(mi)
Problem: Input/Output • Given a set of sensor messages that are generated periodically, their associated deadlines and their routes • Schema of scheduling message transmissions at each hop so that end-to-end deadline violations are minimized
CSMA/CA networks RTS/CTSand false blocking all nodes in the vicinity that receive RTS messages must avoid transmission for the transmission duration, and are thus blocked. So, node 2 will be blocked when m1 is transmitted . node 4 will be blocked when m2 is transmitted.
Example Because of RTS blocking, m2 could not be transmitted concurrently with either m1 or m3
Algorithm • Channel Reuse-based Smallest LST First (global optimize vs. local optimize) • Partition message transmissions into disjoint sets • Messages in one set should be transmitted together • Transmission in one set should complete before next one begins • Once message is scheduled at hop i, it is considered for scheduling at hop i+1 • The scheduler that executes such an algorithm is designed to be a centralized scheduler.
Pseudocode to build transmission set 1.Initially, S←ø; Task set T = {T1, T2, T3 …} is sorted by arrive time and in LST ascending order 2. For each Ti in T do 3. if S = ø 4. build new S1 , let S = S∪ { S1} 5. else 6. for each set Si ∈ S 7. if (feasible(Ti, Si) ) 8. Si = Si ∪ {Ti} 9. end for 10. build new Sj = {Ti}, let S = S ∪{Si} 11.End for • Feasible (Ti, Si) • If arrive time of Ti >= finish time of Si Return FALSE • If finish time of Ti > effective deadline of Ti Return FALSE • If Ti interferes with any Tj, which Tj ∈ Si Return FALSE • If inserting Ti into Si make some messages in the subsequent Sj, i<j<=n, violate the deadline Return FALSE • Return TRUE
Example • S1 = { m1} S = {S1} • m2 interferes with m1 • make new set S2 = {m2} • 3. Check feasible(m3, S1) • s(m3)= 1< f(S1)=2; • f(m3) = max(s(S1),a(m3))+e(m3) = 1+2 =3 • < ed(m3) = 8 • no interfere • extension of f(S1) = 3 doesn’t influence S2 • 4. S1 = {m1, m3} S2 = {m2} • S1 = { m1} S = {S1} • m2 interferes with m1 • make new set S2 = {m2} • 3. Check feasible(m3, S1) • s(m3)= 1< f(S1)=2; • f(m3) = max(s(S1),a(m3))+e(m3) = 1+2 =3 • < ed(m3) = 8 • no interfere • extension of f(S1) = 3 influence S2 • 4. S1 = {m1} S2 = {m2} S3= {m3}
Conclusion • Advantages • Explicitly avoids collisions • Exploits spatial channel reuse • end-to-end effective deadlines can be met as many as possible • Constraints: