Routing Techniques in Wireless Sensor Networks A Survey

  • Slides: 58
Download presentation
Routing Techniques in Wireless Sensor Networks: A Survey J. Al-Karaki, A. E. Kamal A

Routing Techniques in Wireless Sensor Networks: A Survey J. Al-Karaki, A. E. Kamal A Survey on Routing Protocols for Wireless Sensor Networks K. Akkaya, M. Younis

Routing challenges and design issues Ø Node deployment v. Manual deployment q. Sensors are

Routing challenges and design issues Ø Node deployment v. Manual deployment q. Sensors are manually deployed q. Data is routed through predetermined path v. Random deployment q. Optimal clustering is necessary to allow connectivity & energy-efficiency q. Multi-hop routing

Routing challenges and design issues Ø Data routing methods v. Application-specific v. Time-driven: Periodic

Routing challenges and design issues Ø Data routing methods v. Application-specific v. Time-driven: Periodic monitoring v. Event-driven: Respond to sudden changes v. Query-driven: Respond to queries v. Hybrid

Routing challenges and design issues Ø Node/link heterogeneity v. Homogeneous sensors v. Heterogeneous nodes

Routing challenges and design issues Ø Node/link heterogeneity v. Homogeneous sensors v. Heterogeneous nodes with different roles & capabilities q. Diverse modalities q. If cluster heads may have more energy & computational capability, they take care of transmissions to the base station (BS)

Routing challenges and design issues Ø Fault tolerance v. Some sensors may fail due

Routing challenges and design issues Ø Fault tolerance v. Some sensors may fail due to lack of power, physical damage, or environmental interference v. Adjust transmission power, change sensing rate, reroute packets through regions with more power

Routing challenges and design issues Ø Network dynamics v. Mobile nodes v. Mobile events,

Routing challenges and design issues Ø Network dynamics v. Mobile nodes v. Mobile events, e. g. , target tracking v. If WSN is to sense a fixed event, networks can work in a reactive manner q A lot of applications require periodic reporting

Routing challenges and design issues Ø Transmission media v. Wireless channel v. Limited bandwidth:

Routing challenges and design issues Ø Transmission media v. Wireless channel v. Limited bandwidth: 1 – 100 Kbps v. MAC q. Contention-free, e. g. , TDMA or CDMA q. Contention-based, e. g. , CSMA/CA

Routing challenges and design issues Ø Connectivity v. High density high connectivity v. Some

Routing challenges and design issues Ø Connectivity v. High density high connectivity v. Some sensors may die after consuming their battery power v. Connectivity depends on possibly random deployment

Routing challenges and design issues Ø Coverage v. An individual sensor’s view is limited

Routing challenges and design issues Ø Coverage v. An individual sensor’s view is limited v. Area coverage is an important design factor Ø Data aggregation Ø Quality of Service v. Bounded delay v. Energy efficiency for longer network lifetime

Routing Protocols in WSNs Ø I. Flat Ø II. Hierarchical Ø III. Location-based Ø

Routing Protocols in WSNs Ø I. Flat Ø II. Hierarchical Ø III. Location-based Ø IV. Qo. S-based

I. Flat routing

I. Flat routing

Ø Flooding v. Too much waste v. Implosion & overlap v. Channel contention and

Ø Flooding v. Too much waste v. Implosion & overlap v. Channel contention and collisions v. Use in a limited scope, if necessary Ø Data-centric routing v. No globally unique ID v. Naming based on data attributes v. SPIN, Directed diffusion, . . .

SPIN (Sensor Protocols for Information via Negotiation)

SPIN (Sensor Protocols for Information via Negotiation)

SPIN Ø Pros v. Each node only needs to know its one-hop neighbors v.

SPIN Ø Pros v. Each node only needs to know its one-hop neighbors v. Significantly reduce energy consumption compared to flooding Ø Cons v. Data advertisement cannot guarantee the delivery of data q. If the node interested in some data are far from the source and intermediate nodes are not interested in the data, data will not be delivered q. Not good for applications requiring reliable data delivery, e. g. , intrusion detection

Direct Diffusion: Motivation Ø Properties of Sensor Networks v. Data centric v. No central

Direct Diffusion: Motivation Ø Properties of Sensor Networks v. Data centric v. No central authority v. Resource constrained v. Nodes may not know the topology v. Nodes are generally stationary Ø How can we get data from the sensors?

Directed Diffusion: Main Features Ø Data centric v. Individual nodes are unimportant Ø Request

Directed Diffusion: Main Features Ø Data centric v. Individual nodes are unimportant Ø Request driven v. Sink places a request as interest v. Sources satisfying the interest can be found v. Intermediate nodes route data toward sink Ø Localized repair and reinforcement Ø Multi-path delivery for multiple sources, sinks, and queries

Directed Diffusion: Motivating Example Ø Sensor nodes are monitoring animals Ø Users are interested

Directed Diffusion: Motivating Example Ø Sensor nodes are monitoring animals Ø Users are interested in receiving data for all 4 legged creatures seen in a rectangle Ø Users specify the data rate

Directed Diffusion: Interest and Event Naming Ø Query/interest: 1. 2. 3. 4. Type=four-legged animal

Directed Diffusion: Interest and Event Naming Ø Query/interest: 1. 2. 3. 4. Type=four-legged animal Interval=20 ms (event data rate) Duration=10 seconds (time to cache) Rect=[-100, 200, 400] 1. 2. 3. 4. 5. 6. Type=four-legged animal Instance = elephant Location = [125, 220] Intensity = 0. 6 Confidence = 0. 85 Timestamp = 01: 20: 40 Ø Reply: Ø Attribute-Value pairs, no advanced naming scheme

Directed Diffusion: Interest Propagation Ø Flood interest Ø Constrained or Directional flooding based on

Directed Diffusion: Interest Propagation Ø Flood interest Ø Constrained or Directional flooding based on location is possible Ø Directional propagation based on previously cached data Gradient Source Interest Sink

Directed Diffusion: Data Propagation Ø Multipath routing v. Consider each gradient’s link quality Gradient

Directed Diffusion: Data Propagation Ø Multipath routing v. Consider each gradient’s link quality Gradient Source Data Sink

Directed Diffusion: Reinforcement Ø Reinforce one of the neighbor after receiving initial data. v

Directed Diffusion: Reinforcement Ø Reinforce one of the neighbor after receiving initial data. v Neighbor who consistently performs better than others v Neighbor from whom most events received Gradient Source Data Reinforcement Sink

Directed Diffusion: Negative Reinforcement Ø Explicitly degrade the path by re-sending interest with lower

Directed Diffusion: Negative Reinforcement Ø Explicitly degrade the path by re-sending interest with lower data rate. Ø Time out: Without periodic reinforcement, a gradient will be torn down Gradient Source Data Reinforcement Sink

Directed Diffusion: Summary of the protocol

Directed Diffusion: Summary of the protocol

Directed Diffusion: Pros & Cons Ø Different from SPIN in terms of on-demand data

Directed Diffusion: Pros & Cons Ø Different from SPIN in terms of on-demand data querying mechanism v. Sink floods interests only if necessary q. A lot of energy savings v. In SPIN, sensors advertise the availability of data Ø Pros v. Data centric: All communications are neighbor to neighbor with no need for a node addressing mechanism v. Each node can do aggregation & caching

Ø Cons v. On-demand, query-driven: Inappropriate for applications requiring continuous data delivery, e. g.

Ø Cons v. On-demand, query-driven: Inappropriate for applications requiring continuous data delivery, e. g. , environmental monitoring v. Attribute-based naming scheme is application dependent q. For each application it should be defined a priori q. Extra processing overhead at sensor nodes

Extension of Directed Diffusion Ø One-phase pull v. Propagate interest v. A receiving node

Extension of Directed Diffusion Ø One-phase pull v. Propagate interest v. A receiving node pick the link that delivered the interest first v. Assumes the link bidirectionality Ø Push diffusion v. Sink does not flood interest v. Source detecting events disseminate exploratory data across the network v. Sink having corresponding interest reinforces one of the paths

Rumor Routing Ø Variation of directed diffusion v. Don’t flood interests (or queries) v.

Rumor Routing Ø Variation of directed diffusion v. Don’t flood interests (or queries) v. Flood events when the number of events is small but the number of queries large v. Route a query to the nodes that have observed a particular event v. Long-lived packets, called agents, flood events through the network v. When a node detects an event, it adds the event to its events table, and generates an agent v. Agents travel the network to propagate info about local events q. An agent is associated with TTL (Time-To-Live)

Rumor Routing Ø When a node generates a query, a node knowing the route

Rumor Routing Ø When a node generates a query, a node knowing the route to a corresponding event can respond by looking up its events table v. No need for query flooding v. Only one path between the source and sink v. Rumor routing works well only when the number of events is small v. Cost of maintaining a large number of agents and large event tables will be prohibitive v. Heuristic for defining the route of an event agent highly affects the performance of next-hop selection

MCFA (Minimum Cost Forwarding Algorithm) Ø Assume the direction of routing is always known,

MCFA (Minimum Cost Forwarding Algorithm) Ø Assume the direction of routing is always known, i. e. , toward the fixed base station (BS) Ø No need for a node to have a unique ID or routing table Ø Each node maintains the least cost estimate from itself to BS Ø Broadcast a message to neighbors Ø A neighbor checks if it’s on the least cost path btwn the source and BS Ø If so, it re-broadcasts the message to its neighbors Ø Repeat until BS is reached

MCFA Ø Each node has to know the least cost path estimate to BS

MCFA Ø Each node has to know the least cost path estimate to BS v. BS broadcasts a message with cost set to 0 v. Every node initially sets its cost to BS to ∞ v. When a node receives the msg from BS, it checks if the estimate in the packet + 1 < the node’s current estimate to BS q. If yes, the current estimate & estimate in the msg are updated and resent q. Else, delete the msg; Do nothing v. A node far from BS may receive several msg’s A node will not send the updated msg until constant k * link cost Ø Works well for fixed topologies

Gradient-Based Routing (GBR) Ø Variation of directed diffusion Ø Each node memorizes the number

Gradient-Based Routing (GBR) Ø Variation of directed diffusion Ø Each node memorizes the number of hops when the interest is diffused Ø Each node computes its height, i. e. , the minimum number of hops to BS Ø Difference btwn a node’s height & its neighbor’s is the gradient on the link Ø Forward a packet on a link with the largest gradient Ø Data aggregation v. When multiple paths pass through a node, the node can combine data

Ø Traffic spreading v. Uniformly divide traffic over the network to increase network lifetime

Ø Traffic spreading v. Uniformly divide traffic over the network to increase network lifetime q. Stochastic scheme: Randomly pick a gradient when two or more next hops have the same gradient q. Energy-based scheme: A node increases its height when its energy drops below a certain threshold q. Stream-based scheme: New streams are not routed through nodes that are part of the path for other streams Ø Outperforms directed diffusion in terms of total energy

COUGAR & Tiny. DB Ø View a WSN as a distributed database Ø Use

COUGAR & Tiny. DB Ø View a WSN as a distributed database Ø Use declarative queries to abstract query processing from the network layer—network layer independent Ø Perform in-network data aggregation Ø Drawbacks v. Extra overhead & energy consumption due to the extra query layer v. Synchronization is required for data aggregations v. Leader nodes should be dynamically maintained to prevent them from being hotspots

ACQUIRE Ø View a WSN as a distributed DB Ø Complex queries can be

ACQUIRE Ø View a WSN as a distributed DB Ø Complex queries can be divided into subqueries Ø BS sends a query Ø Each node tries to answer the query by using precached info and forwards the query to another node Ø If the cached info is not fresh, the nodes gather info from their neighbors within a lookahead of d hops Ø Once the query is resolved completely, it is sent back to BS via the reverse path or shortest path

Ø ACQUIRE can deal with complex queries by allowing many nodes to send responses

Ø ACQUIRE can deal with complex queries by allowing many nodes to send responses v. Directed diffusion cannot handle complex queries due to too much flooding v. ACQUIRE can adjust d for efficient query processing v. If d = network diameter, ACQUIRE becomes similar to flooding v. On the other hand, a query has to travel more if d is too small v. Provides mathematical modeling to find an optimal value of d for a grid of sensors, but no experiments performed

II. Hierarchical Routing

II. Hierarchical Routing

LEACH (Low Energy Clustering Hierarchy) Ø Cluster-based protocol Ø Each node randomly decides to

LEACH (Low Energy Clustering Hierarchy) Ø Cluster-based protocol Ø Each node randomly decides to become a cluster heads (CH) Ø CH chooses the code to be used in its cluster v CDMA between clusters Ø CH broadcasts Adv v Each node decides to which cluster it belongs based on the received signal strength of Adv Ø CH creates a transmission schedule for TDMA in the cluster Ø Nodes can sleep when its not their turn to transmit Ø CH compresses data received from the nodes in the cluster and sends the aggregated data to BS Ø CH is rotated randomly

LEACH v. Pros q. Distributed, no global knowledge required q. Energy saving due to

LEACH v. Pros q. Distributed, no global knowledge required q. Energy saving due to aggregation by CHs v. Shortcomings q. LEACH assumes all nodes can transmit with enough power to reach BS if necessary (e. g. , elected as CHs) q. Each node should support both TDMA & CDMA v. Extension of LEACH [5] q. High level negotiation, similar to SPIN q. Only data providing new info is transmitted to BS

Comparison between SPIN, LEACH & Directed Diffusion SPIN Optimal No Route Network Good Lifetime

Comparison between SPIN, LEACH & Directed Diffusion SPIN Optimal No Route Network Good Lifetime Resource Yes Awareness LEACH No Directed Diffusion Yes Very good Good Yes (? )

TEEN (Threshold sensitive Energy Efficient Network protocol) Ø Reactive, event-driven protocol for time-critical applications

TEEN (Threshold sensitive Energy Efficient Network protocol) Ø Reactive, event-driven protocol for time-critical applications v A node senses the environment continuously, but turns radio on and xmit only if the sensor value changes drastically v No periodic xmission q Don’t wait until the next period to xmit critical data q Save energy if data is not critical Ø CH sends its members a hard & a soft threshold v Hard threshold: A member only sends data to CH only if data values are in the range of interest v Soft threshold: A member only sends data if its value changes by at least the soft threshold v Every node in a cluster takes turns to become the CH for a time interval called cluster period Ø Hierarchical clustering

Multi-level hierarchical clustering in TEEN & APTEEN

Multi-level hierarchical clustering in TEEN & APTEEN

TEEN Ø Good for time-critical applications Ø Energy saving v. Less energy than proactive

TEEN Ø Good for time-critical applications Ø Energy saving v. Less energy than proactive approaches v. Soft threshold can be adapted v. Hard threshold could also be adapted depending on applications Ø Inappropriate for periodic monitoring, e. g. , habitat monitoring Ø Ambiguity between packet loss and unimportant data indicating no drastic change

APTEEN (Adaptive Threshold sensitive Energy Efficient Network protocol) Ø Extends TEEN to support both

APTEEN (Adaptive Threshold sensitive Energy Efficient Network protocol) Ø Extends TEEN to support both periodic sensing & reaction to time critical events Ø Unlike TEEN, a node must sample & transmit a data if it has not sent data for a time period equal to CT (count time) specified by CH Ø Compared to LEACH, TEEN & APTEEN consumes less energy (TEEN consumes the least) v Network lifetime: TEEN ≥ APTEEN ≥ LEACH Ø Drawbacks of TEEN & APTEEN v Overhead & complexity of forming clusters in multiple levels and implementing threshold-based functions

Sensor aggregate routing Ø Sensor aggregate: a set of nodes satisfying a grouping predicate

Sensor aggregate routing Ø Sensor aggregate: a set of nodes satisfying a grouping predicate Ø Mainly designed for target tracking Source: M. Handy at University of Rostock

TTDD (Two Tier Data Dissemination) Ø Data dissemination to mobile sinks Ø Two-tier query

TTDD (Two Tier Data Dissemination) Ø Data dissemination to mobile sinks Ø Two-tier query & data forwarding Ø Objectives v. Source proactively builds a grid structure to support data availability for mobile sinks q. Mobility pattern is unknown a priori v. Localize impacts of sink mobility on data forwarding v. Only a small set of sensor nodes maintain forwarding state

TTDD: Sensor Network Model Sink Stimulus Source: TTDD at Mobicom ‘ 02 Sink

TTDD: Sensor Network Model Sink Stimulus Source: TTDD at Mobicom ‘ 02 Sink

TTDD Basics Dissemination Node Data Announcement Source Data Sink Immediate Query Dissemination Node Source:

TTDD Basics Dissemination Node Data Announcement Source Data Sink Immediate Query Dissemination Node Source: TTDD at Mobicom ‘ 02

TTDD Mobile Sinks Dissemination Node Trajectory Forwarding Data Announcement Source Immediate Dissemination Node Data

TTDD Mobile Sinks Dissemination Node Trajectory Forwarding Data Announcement Source Immediate Dissemination Node Data Sink Immediate Dissemination Node Source: TTDD at Mobicom ‘ 02 Trajectory Forwarding

TTDD Multiple Mobile Sinks Dissemination Node Trajectory Forwarding Data Announcement Source Immediate Dissemination Node

TTDD Multiple Mobile Sinks Dissemination Node Trajectory Forwarding Data Announcement Source Immediate Dissemination Node Data Source: TTDD at Mobicom ‘ 02

Grid Maintenance Dissemination Node Source X Data Source: TTDD at Mobicom ‘ 02 Immediate

Grid Maintenance Dissemination Node Source X Data Source: TTDD at Mobicom ‘ 02 Immediate Dissemination Node

III. Location-based routing protocols

III. Location-based routing protocols

GAF (Geographic Adaptive Fidelity) Ø Energy-aware location-based protocol mainly designed for MANET Ø Each

GAF (Geographic Adaptive Fidelity) Ø Energy-aware location-based protocol mainly designed for MANET Ø Each node knows its location via GPS v Associate itself with a point in the virtual grid v Nodes associated with the same point on the grid are considered equivalent in terms of the cost of packet routing v Node 1 can reach any of nodes 2, 3 & 4 2, 3, 4 are equivalent; Any of the two can sleep without affecting routing fidelity

GAF Ø Three states v. Discovery: Determine neighbors in a grid v. Active v.

GAF Ø Three states v. Discovery: Determine neighbors in a grid v. Active v. Sleep Ø Each node in the grid estimates its time of leaving the grid and sends it to its neighbors v. The sleeping neighbors adjust their sleeping time to keep the routing fidelity

GEAR (Geographic and Energy Aware Routing) Ø Restrict the number of interest floods in

GEAR (Geographic and Energy Aware Routing) Ø Restrict the number of interest floods in directed diffusion v. Consider only a certain region of the network rather than flooding the entire network Ø Estimated cost = f(residual energy, distance to the destination) Ø Learned cost is propagated one hop back every time a packet reaches the sink v. Route setup for the next packet can be adjusted

GEAR Ø Phase 1: Forwarding packets towards the region v. Forward a packet to

GEAR Ø Phase 1: Forwarding packets towards the region v. Forward a packet to the neighbor minimizing the cost function f q. Forward data to the neighbor which is closest to the sink and has the highest level of remaining energy v. If all neighbors are farther than itself, there is a hole Pick one of the neighbors based on the learned cost

GEAR Ø Phase 2: Forwarding the packet within the target region v. Apply recursive

GEAR Ø Phase 2: Forwarding the packet within the target region v. Apply recursive forwarding q. Divide the region into four subareas and send four copies of the packet q. Repeat this until regions with only one node are left v. Alternatively apply restricted flooding q. Apply when the node density is low Ø GEAR successfully delivers significantly more packets than GPSR (Greedy Perimeter Stateless Routing) v. GPSR will be covered in detail in another class

Summary

Summary

Questions?

Questions?