Survey on Routing in Sensor Networks Lichun Bao

Survey on Routing in Sensor Networks Lichun Bao Computer Science Department University of California, Irvine

Sensor Node Components n n n Sensing Processing Transmission Mobilizer position finding system Power units

Purposes of Sensor Networks n n Information processing: distributed computing, target field imaging, weather monitoring Detection: q q n intrusion detection, security and tactical surveillance detecting ambient conditions such as temperature, movement, sound, light, or the presence of certain objects, inventory control Disaster management.

Components of Routing n n n Control function: energy-efficient route discovery Data-forwarding functions: relaying of data from the sensor nodes to the BS Addressing scheme q q Content-based Topology-based n q Subnetting Location-based

Address Centric Routing (AC) Temperature Reading (source 2) Temperature Reading (source 1) source 2 source 1 source 2 Z B source 1 source 2 Give Me The Average Temperature? ( sink )

Data Centric Routing (DC) Temperature Reading (source 2) Temperature Reading (source 1) source 2 source 1 source 2 Z B source 1 & 2 Give Me The Average Temperature? ( sink )

Requirements on Routing n n n Self-organizing Tailor toward specific traffic pattern (BScentric, application specific, position-aware) Constrained in energy, processing, storage Stationary deployment Redundancy – in-network processing

Classification of Routing Protocols n Networking structure q q q n Flat Hierarchical Location-based Network operation q Data-reporting model: time-driven (continuous), eventdriven, query-driven, and hybrid n q q a diverse mixture of sensors for monitoring temperature, pressure and humidity of the surrounding environment, detecting motion via acoustic signatures, and capturing the image or video tracking of moving objects. multipath-based negotiation-based Qo. S-based Coherent-based

Classification

Design Considerations n n n n n Node deployment Energy consumption Data Reporting Model Node/Link Heterogeneity Fault Tolerance Scalability Network dynamics Coverage Qo. S

Overview n What is Data Routing? q n What is Data Aggregation? q n Optimal Aggregation Sensor Query Engines q n Directed Diffusion [Estrin, 2000] Iris. Net [Nath, 2003] Using XML in Sensor Networks

Basic Mechanisms of SPIN : Flat Routing n Sensor Protocol for Information via Negotiation q q q Negotiation before exchanging data Negotiation via Meta-Data Resource-aware and resource-adaptive

SPIN Protocols n n SPIN-PP: a 3 -stage handshake protocol for Point-to-Point media SPIN-EC: SPIN-PP with a low-energy threshold for Energy Conservation. SPIN-BC: a 3 -stage handshake protocol for Broad. Cast media SPIN-RL: SPIN-BC for lossy networks

How does SPIN-PP work? D A B DATA message E ADV message C REQ message

SPIN-EC n n SPIN-PP with a simple energy-conservation heuristic When the low-energy threshold is observed, the node reduces its participation in the protocol.

SPIN-BC B E ADV B ADV A REQ A C REQ D ADV C D (1) E E B DATA C DATA A DATA D (3) (2)

SPIN-RL n n Re-request data from the original node or a random neighbor node After sending out a DATA message, the node will wait for a predetermined amount of time before responding to any more request for that piece of data

Performance n n In a point-to-point network, SPIN can deliver 60% more data for a given amount of energy than conventional approaches. In a broadcast network, SPIN delivers 80% more data for a given amount of energy than conventional approaches.

Directed Diffusion (interests propagation) Temperature A Temperature B A B Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: E C D E Interest Cache Sink Type F Temperatur e Sound Z Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: D Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: F F Give Me The Temperature? ( sink ) Interest Propagation Phase

Directed Diffusion (gradient Interest Type: Temperature establishment) Interval: 1 s Temperature A Temperature B A B Data Type: Temperature Location: A Value: 58 Time: 00: 03 Data Type: Temperature Location: B Value: 61 Time: 00: 03 Duration: 10 m Sink: C C D E Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: D Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: F F Give Me The Temperature? Interest & Data Propagation Phase ( sink )

Gradient Table - Temperature Value Node Expires 1 s E 10 m 2 s Z 1 m Directed Diffusion Temperature A Temperature B A B Gradient Type: Temperature Interval: 1 s Duration: 10 m Data Type: Temperature Location: B Value: 61 Time: 00: 04 C D E Data Type: Temperature Location: A, B Value: 57, 61 Time: 00: 03, 00: 03 Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: D Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: F F Give Me The Temperature? Interest & Data Propagation Phase ( sink )

Gradient Table - Temperature Value Node Expires 1 s E 10 m 1 s C 10 m 2 s Z 1 m Temperature A Temperature B A B Gradient Type: Temperature Interval: 1 s Duration: 10 m Interest – Reinforce Type: Temperature Interval: 10 ms Duration: 10 m Sink: F Recipient: E Gradient Type: Temperature Interval: 1 s Duration: 10 m E Gradient Table - Temperature Value Node Expires 10 ms F 10 m 2 s P 1 m Gradient Type: Temperature Interval: 10 ms Duration: 10 m Gradient Type: Temperature Interval: 1 s Duration: 10 m C D Data Type: Temperature Location: B Value: 61 Time: 00: 04 Interest Type: Temperature Interval: 1 s Duration: 10 m Sink: F F Give Me The Temperature? Interest & Data Propagation Phase ( sink )

Gradient Table - Temperature Value Node Expires 10 ms E 10 m 2 s Z 1 m Value Node Expires 10 ms E 10 m 1 s C 10 m 2 s Z 1 m Temperature A Temperature B A B Gradient Type: Temperature Interval: 10 ms Duration: 10 m Interest – Reinforce Type: Temperature Interval: 10 ms Duration: 10 m Sink: E Recipient: A, B Gradient Type: Temperature Interval: 10 ms Duration: 10 m C Gradient Type: Temperature Interval: 1 s Duration: 10 m D E Gradient Type: Temperature Interval: 10 ms Duration: 10 m Gradient Type: Temperature Interval: 1 s Duration: 10 m Data Type: Temperature Location: B Value: 61 Time: 00: 04 F Give Me The Temperature? Interest & Data Propagation Phase ( sink )

Temperature A Temperature B A B Gradient Type: Temperature Interval: 10 s Duration: 10 m E Gradient Type: Temperature Interval: 1 s Duration: 10 m C Gradient Type: Temperature Interval: 1 s Duration: 10 m D Interest – Negative Type: Temperature Interval: 5 s Duration: 10 m Gradient Sink: F Gradient Negative Type: Temperature Recipient: D Interval: 10 ms Interval: 5 s Duration: 10 m F Give Me The Temperature? Interest & Data Propagation Phase ( sink )

Gradient Table - Temperature Value Node Expires 10 ms E 10 m 5 s C 10 m 2 s Z 1 m Temperature A Temperature B A B Gradient Type: Temperature Interval: 10 s Duration: 10 m E Gradient Type: Temperature Interval: 5 s Duration: 10 m C Gradient Type: Temperature Interval: 5 s Duration: 10 m D Interest – Negative Type: Temperature Interval: 5 s Duration: 10 m Gradient Sink: F Gradient Type: Temperature Recipient: D Interval: 10 ms Interval: 5 s Duration: 10 m F Give Me The Temperature? Interest & Data Propagation Phase ( sink )

Directed Diffusion n Data-centric Routing Quality of Data influences the network Application Specific q What to reinforce? n q Gradients could be custom n n Is it time to delivery? Is it energy conservation? Could be a threshold of when to send data Robust to network changes q q Multiple paths are kept Reinforce those paths if necessary

Gradient Table - Temperature Value Node Expires 10 ms E DEAD 10 ms C 10 m 2 s Z 1 m Temperature A Temperature B A B X X X Gradient Type: Temperature Interval: 10 s Duration: 10 m Gradient Type: Temperature Interval: 5 s Duration: 10 m C Gradient Type: Temperature Interval: 5 s Duration: 10 m D E Interest – Reinforce Type: Temperature Interval: 10 ms Duration: 10 m Sink: F While waiting for Gradient data along the Type: Temperature Recipient: D broken path node F Interval: 5 s times out Duration: 10 m F Give Me The Temperature? ( sink ) Recovering from broken links X

Data Aggregation Temperature Reading (source 2) Temperature Reading (source 1) source 2 source 1 source 2 Aggregates the data before routing it source 1 & 2 Give Me The Average Temperature? ( sink ) In this example average would aggregate to: <sum, count>

Rumor Routing n n Nodes having observed an event send out agents which leave routing info to the event as state in nodes Agents attempt to travel in a straight line If an agent crosses a path to another event, it begins to build the path to both Agent also optimizes paths if they find shorter ones. Paper: David Braginsky and Deborah Estrin. Slide adapted from Sugata Hazarika, UCLA

Algorithm Basics n n All nodes maintain a neighbor list. Nodes also maintain a event table q n When it observes an event, the event is added with distance 0. Agents q q Packets that carry local event info across the network. Aggregate events as they go.

Agent Path n Agent tries to travel in a “somewhat” straight path. q q q Maintains a list of recently seen nodes. When it arrives at a node adds the node’s neighbors to the list. For the next tries to find a node not in the recently seen list. Avoids loops -important to find a path regardless of “quality”

Following Paths n n A query originates from source, and is forwarded along until it reaches it’s TTL Forwarding Rules: q q q If a node has seen the query before, it is sent to a random neighbor If a node has a route to the event, forward to neighbor along the route Otherwise, forward to random neighbor using straightening algorithm

Minimum Cost Forwarding Algo (MCFA) Gradient-based Routing (GBR) n n Bellman-ford algorithm relaxing the distance from the BS to all the sensors GBR has adjustable height for a node to divert traffic to other nodes.

Information-driven sensor querying (IDSQ) Constrained anisotropic diffusion routing (CADR) n n n Introduction of two novel techniques IDSQ and CADR for energy-efficient data querying and routing in sensor networks the use of general form of information utility that models the information content as well as the spatial configuration of a network generalization of directed diffusion that uses both the communication cost and the information utility to diffuse data

Problem Formulation n n zi (t) = h(x(t), i (t)), (1) x(t) is based on parameters of the sensor, i (t) and zi (t) are characteristics and measurement of sensor i respectively. for sensors measuring sound amplitude i = [ x i , σi 2 ] T (3) xi is the known sensor position and σi 2 is the known additive noise variance zi = a / || xi - x || /2 + wi , (4) a is target amplitude, is attenuation coefficient , wi is Gaussian noise with variance σi 2

Define Belief as. . . n n n representation of the current a posteriori distribution of x given measurement z 1, …, z. N: p(x | z 1, …, z. N) expectation is considered estimate x = xp(x | z 1, …, z. N)dx covariance approximates residual uncertainty = (x - x)Tp(x | z 1, …, z. N)dx

Define Information Utility as … The Information Utility function is defined as Ψ: P(Rd) R d is the dimension of x Ψ assigns value to each element of P(Rd) indicating the uncertainty of the distribution smaller value -> more spread out distribution larger value -> tighter distribution

A lot more probability theory to follow on IDSQ and CADR n Incrementally construct sensor set to follow events n Mystery to pursue …

COUGAR, ACQUIRE n n View the network as a distributed database Complex queries are further divided into several sub queries

Hierarchical Routing n LEACH - Low Energy Adaptive Clustering Hierarchy q q q Use clusterheads to form groups for data aggregation Use TDMA with clusters, and CDMA between clusters. LEACH assumes that all nodes can transmit with enough power to reach the BS if needed and that each node has computational power to support different MAC protocols.

TEEN and APTEEN n n TEEN (Threshold-sensitive Energy Efficient sensor Network protocol) APTEEN (Adaptive Periodic Thresholdsensitive Energy Efficient sensor Network protocol) Hard-threshold for initiating data reporting Soft-threshold for incremental changes

Grid Construction n Virtual Grid Architecture routing (VGA): q q n Assumes nodes are arranged in a fixed topology Select Local Aggregators (LAs) and Master Aggregators (MAs) for data collection Two-Tier Data Dissemination (TTDD) q q q Assumes known sensor locations Grids are formed by fixed distance between CHs Other nodes can be put to sleep to save energy

Data Aggregation n We can reduce the data packet size by aggregating data en route Parallel Process large aggregations of data in the network Scale to large numbers of both sinks and sources

Optimal Data Aggregation n Optimal Data Aggregation is NP-Hard q Sub-optimal Algorithms: n Opportunistic q n Center at Nearest Source (CNS) [1] q n Sources send using shortest path if able aggregate Greedy Incremental Tree (GIT) [1] q n The nearest source acts as the aggregation point Shortest Paths Tree (SPT) [1] q n Just aggregate when possible Recursively select the closest source to the tree Clustered Diffusion w/ Dynamic Data Aggregation [2] q Hybrid between diffusion and clustering with the ability to aggregate data at the cluster heads 1 – Krishnamachari, Estrin, & Wicker, 2002 2 - Chatterjea & Havinga, 2003

Querying Sensor Networks n TAG (Tiny Aggregation) [Madden, 2002] q q n Cornell’s COUGAR [Bonnet, 2000] q q q n Individual sensor data Distributed Gathering SQL-like Niagara. CQ [Chen, 2000] q q n Focus on Aggregation using SQL-like query language Integrated in Tiny. OS Focus on Continuous Queries XML Streams Iris. Net [Nath, 2003] q q Individual sensor data “World-Wide Sensor Network”

Coherent-based category n n In non-coherent data processing routing, nodes will locally process the raw data before being sent to other nodes for further processing. In coherent routing, the data is forwarded to aggregators after minimum processing.

Overall Comparison

Future/current directions n all share the common objective of prolonging the network lifetime. q q q q Exploit redundancy Tiered architectures (mix of form/energy factors) Achieve desired global behavior with adaptive localized algorithms Perform in-network distributed processing Time and location synchronization Localization Secure Routing