Unit 33-routing protocols for wsn
Download as PPTX, PDF0 likes1,192 views
This document discusses wireless sensor networks and routing protocols. It covers several key topics: 1) It describes single-hop and multihop data transmission in wireless sensor networks and the advantages of multihop in increasing network lifetime and reducing interference. 2) It discusses routing challenges in wireless sensor networks due to constraints like energy, bandwidth and changing environments. It also covers routing strategies like proactive, reactive and hybrid routing. 3) It provides details on common routing protocols for wireless sensor networks like flooding, gossiping, SPIN and LEACH, outlining their key mechanisms and advantages/disadvantages. LEACH uses clustering to improve energy efficiency.
Unit 33-routing protocols for wsn
- 1. Wireless Sensor Networks Prepared By: Dr. Nagarathna and Deepika Dept. of CS & E PESCE, Mandya 1
- 2. ROUTING PROTOCOLS FOR WSN 2
- 3. DATA DISSEMINATION AND GATHERING The way that data and queries are forwarded between the base station and the location of the target Different Approaches Single Hop Multihop 3
- 4. Single Hop • Sensor node exchange data directly with the base station – Costly • As nodes that are farther away from the base station may deplete their energy quickly ,thereby limiting the lifetime of the network 4
- 5. 5
- 6. MULTIHOP Data exchange between the sensors and the base stations is usually carried out using multihop packet transmission over short communication radius Significant energy savings( Increase the network life) Reduces communication interference between sensor nodes competing to access the channel Contd…6
- 7. 7
- 8. Multihop WSN • Intermediate nodes must participate in forwarding data packets between the source and the destination • Principal task of the routing – Determining set of intermediate nodes to form a data-forwarding path between the source and the destination • Routing must address multiple challenging design requirements – Correctness – Stability – Optimality with respect to various performance metrics Contd..8
- 9. Multihop WSN • Properties of WSNs combined with constraints – Energy – Bandwidth 9
- 10. ROUTING CHALLENGES AND DESIGN ISSUES IN WIRELESS SENSOR NETWORKS • Challenges can be attributed to multiple factors including – Severe energy constraints – Limited computing – Communication capabilities – Dynamically changing environment – Unique data traffic models – Application-level quality of service 10
- 11. Network Scale and Time-Varying Characteristics • The densities of the WSNs may vary widely, ranging from very sparse to very dense. – The number of sensor nodes from hundreds to thousands are deployed in – Ad hoc (for current situation) – unsupervised manner over wide coverage areas 11
- 12. Network Scale and Time-Varying Characteristics • The behavior of sensor nodes – Dynamic and highly adaptive • To self-organize and conserve energy – sensor nodes adjust their behavior constantly in response to their current level of activity • Sensor nodes adjust their behavior in response to the Erratic and unpredictable behavior of wireless connections – Caused by high noise levels and radio-frequency interference – To prevent severe performance degradation of the application supported. 12
- 13. Resource Constraints • Sensor nodes are designed with minimal complexity for large-scale deployment at a reduced cost • Must achieve a long lifetime operating on limited battery • Multihop packet transmission source of power consumption • Reducing energy consumption by dynamically controlling the duty cycle of the wireless sensors 13
- 14. Sensor Applications Data Models • The Data model – Describes the flow of information between the sensor nodes and the data sink. – Highly dependent on the nature of the application (requested and used) – Several data models have been proposed to • Address the data-gathering needs • Interaction requirements of a variety of sensor applications 14
- 15. ROUTING STRATEGIES IN WIRELESS SENSOR NETWORKS • Classic trade-off between responsiveness and efficiency • Accommodate the limited processing and communication capabilities of nodes • overhead for mobile nodes is measured in terms of – Bandwidth utilization – Power consumption – Processing requirements 15
- 16. Routing algorithms for ad hoc networks • Classified according to the manner – How Information is acquired and maintained – How Information is used to compute paths • Three different strategies identified: – Proactive – Reactive – Hybrid 16
- 17. Proactive Strategy • Also referred as table driven • Periodic dissemination of routing information • Maintain consistent and accurate routing tables across all nodes of the network. • The structure of the network can be either – Flat – Hierarchical 17
- 18. Proactive Strategy • Flat proactive routing – strategies needs to compute optimal paths – The overhead required to compute these paths in a dynamically changing environment • Hierarchical routing – Better suited to meet the routing demands of large ad hoc networks. 18
- 19. Reactive routing strategies Establish routes to a limited set of destinations on demand Do not maintain global information across all nodes of the network Rely on a dynamic route search to establish paths between a source and a destination Typically involves flooding a route discovery query, with the replies traveling back along the reverse path. Strategies vary in the way Control the flooding process to reduce communication overhead The way routes are computed and reestablished when failure occurs 19
- 20. Hybrid strategies • To achieve stability and scalability in large networks • Network is organized into mutually adjacent clusters maintained dynamically as nodes join and leave their assigned clusters. • A hybrid routing strategy can be adopted whereby – Proactive routing is used within a cluster – Reactive routing is used across clusters. • The main challenge is to reduce the overhead required to maintain the clusters. 20
- 21. Design Requirements • The design of routing protocols for WSNs must consider – Power and resource limitations of the network nodes – Time-varying quality of the wireless channel – Possibility for packet loss and delay 21
- 22. WSN Routing Techniques • Flat network architecture • Clusters Based • Data-centric approach • Location based 22
- 23. WSN Routing Techniques • Flat network architecture : Nodes are considered peers. • Advantages – Minimal overhead to maintain the infrastructure – Potential for the discovery of multiple routes between communicating nodes for fault tolerance 23
- 24. WSN Routing Techniques • Clusters Based (Hierarchical Routing) – Achieve energy efficiency, stability, and scalability. – A node with higher residual energy will cluster head – The cluster head is responsible for coordinating activities within the cluster and forwarding information between clusters • Clustering has potential to reduce energy consumption and extend the lifetime of the network24
- 25. WSN Routing Techniques Data-centric approach • Approach to disseminate interest within the network • A source node queries an attribute for the phenomenon rather than an individual sensor node • The interest dissemination is achieved by assigning tasks to sensor nodes and expressing queries to relative to specific attributes • Different strategies can be used to communicate interests to the sensor nodes – Broadcasting – Attribute based multicasting – Geo-casting – Anycasting 25
- 26. WSN Routing Techniques Location based • Protocols uses location to address a sensor node. • Location- based routing is useful – In applications where the position of the node within the geographical coverage of the network is relevant to the query issued by the source node • Query may specify a specific – Area where a phenomenon of interest may occur or – The vicinity to a specific point in the network environment 26
- 27. Flooding and Its Variants • Flooding – Uses a reactive approach whereby each node on receiving a data or control packet sends the packet to all its neighbors. – After transmission, a packet follows all possible paths. – Unless the network is disconnected, the packet will eventually reach its destination – If network topology changes, the packet transmitted follows the new routes. Contd…27
- 28. Contd…28
- 29. Flooding • Disadvantages : – Flooding cause packets to be replicated indefinitely by network nodes • Packet circulates indefinitely in the network • Solution – Hop count Field • Field is included in the packet with value set to approximately the diameter of the network. • As the packet travels across the network, the hop count is decremented by one. • When the hop count reaches zero, the packet is simply discarded. Contd…29
- 30. Flooding • Other Solution – Using a time-to-live field • To records the number of time units that a packet is allowed to live within the network • At the expiration of this time, the packet is no longer forwarded. – Drop all the packets that it has already forwarded • Requires maintaining at least a recent history of the traffic, to keep track of which data packets have already been forwarded. 30
- 31. Drawback of flooding 1. Traffic implosion Effect is caused by duplicate control or data packets being sent repeatedly to the same node. 31
- 32. Drawback of flooding 2. Overlapping problem occurs when two nodes covering the same region send packets containing similar information to the same node 3. Resource blindness : The forwarding rule used to route packets does not take into consideration the energy constraints of the sensor nodes. Node’s energy may deplete rapidly, reducing considerably the lifetime of the network. 32
- 33. Gossiping • Address the shortcomings of flooding • Derivative approach referred to as gossiping • Each node sends the incoming packet to a randomly selected neighbor • The neighbor selected randomly chooses one of its own neighbors and forwards the packet to the neighbor chosen. • The process continues iteratively – Until the packet reaches its intended destination OR – The maximum hop count is exceeded 33
- 34. Gossiping • Advantages – Gossiping avoids the implosion problem by limiting the number of packets that each node sends to its neighbor to one copy. • Disadvantages – The latency that a packet suffers on its way to the destination may be excessive, particularly in a large network. – Caused primarily by the random nature of the protocol, which, in essence, explores one path at a time. 34
- 35. Sensor Protocols for Information via Negotiation (SPIN) • Data-centric negotiation based information dissemination protocols • Objective – To efficiently disseminate observations gathered by individual sensor nodes to all the sensor nodes in the network 35
- 36. SPIN The basic of protocols are Data negotiation Resource adaptation Nodes running SPIN ‘‘learn’’ about the content of the data before any data are transmitted between network nodes Nodes associate metadata with data they produce to perform negotiations before transmitting the actual data (eliminates the possibility of overlap) 36
- 37. SPIN A receiver that expresses interest in the data content can send a request to obtain the data advertised Assures Data are sent only to interested nodes Eliminating traffic implosion Reducing significantly the transmission of redundant data throughout the network 37
- 38. 38 SPIN • To carry out negotiation and data transmission, nodes running SPIN use three types of messages – ADV (advertisement) – REQ (request) – DATA • ADV: – Used to advertise new data among nodes – A network node that has data to share can advertise its data – Transmitting an ADV message containing the metadata describing the data.
- 39. SPIN • REQ: – Used to request an advertised data of interest – Upon receiving an ADV containing metadata a network node interested in receiving specific data sends a REQ message to the metadata advertising node • DATA: – Contains the actual data collected by a sensor, along with a metadata header – Message is typically larger than the ADV and REQ messages. 39
- 40. 40 SPIN basic protocol operations
- 41. SPIN-PP The simplest version of SPIN referred to as point –to- point comm. network. The three-step handshake protocol used by SPIN-PP 1. The node holding the data, node A, issues an advertisement packet (ADV) 2. Node B expresses interest in receiving the data by issuing a data request (REQ) 3. Node A responds to the request and sends a data packet to node B (DATA) 41
- 42. SPIN-PP • SPIN-PP uses negotiation to overcome the implosion and overlap • A simulation-based performance study of SPIN-1 shows that the protocol reduces energy consumption by a factor of 3.5 compared to flooding. • The protocol also achieves high data dissemination rates 42
- 43. SPIN-EC SPIN-EC designed for point-to-point communication Incorporates a threshold based resource-awareness mechanism to complete data negotiation If a nodes energy level approaches the low threshold, a node running SPIN-EC reduces its participation in the protocol operations. 43
- 44. SPIN-EC • if a node receives an advertisement, it does not send out an REQ message if its energy resource is not high enough to transmit an REQ message and to receive data. • The simulation results show that SPIN-EC – Disseminates 60% more data per unit energy than flooding – Close to the ideal amount of data that can be disseminated per unit energy. 44
- 45. SPIN-BC Designed for broadcast networks The packet transmitted by a node is received by all the other nodes within a certain range of the sending node Node which has received an ADV message does not respond immediately with an REQ message 45
- 46. SPIN-BC • Node waits for a certain amount of time and – Monitors the communications channel. – If the node hears an REQ message issued by another node which is interested in receiving the data, it cancels its own request(eliminating any redundant requests for the same message) • The advertising node sends the data message only once, even when it receives multiple requests for the same message 46
- 47. SPIN-BC protocol basic operations 47 A sends a ADV packet to advertise the data to its neighbors All nodes hear the advertisement, but node C is first to issue a REQ packet (E & F not interested) Nodes B and D hear the broadcast request and refrain from issuing their own REQ packets Upon hearing node C’s request, node A replies by sending the data packet. All nodes within the transmission range of A receive the data packet, including nodes E and
- 48. SPIN-RL Extends the capabilities of SPIN-BC to enhance Reliability Overcome message transmission errors by a lossy channel. Reliability is achieved by periodic broadcasting of ADV and REQ messages Each node in SPIN-BC keeps track of the advertisements it hears and the nodes where these advertisements originate. If a node requesting specific data of interest does not receive the data requested within a certain period of time, it sends the request again. 48
- 49. SPIN-RL Improved reliability by re-advertising metadata periodically SPIN-RL nodes limit the frequency of resend the data messages After sending out a data message, a node waits for a certain time period before it responds to other requests for the same data message. 49
- 50. Advantages of SPIN • The SPIN protocol family addresses the major drawbacks of flooding and gossiping. • Simulation results show that – SPIN is more energy efficient than flooding or gossiping – SPIN disseminates data greater than or equal to the rate of either of these protocols • Achieves these gains by – localizing topology changes – Eliminating dissemination of redundant information through semantic negotiation 50
- 51. 51 Drawback of SPIN • SPIN’s data advertisement mechanism cannot guarantee delivery of data.
- 52. Low-Energy Adaptive Clustering Hierarchy • Low-energy adaptive clustering hierarchy (LEACH) is a routing algorithm designed to collect and deliver data to a base station • The main objectives of LEACH are: – Extension of the network lifetime – Reduced energy consumption by each network sensor node – Use of data aggregation to reduce the number of communication messages 52
- 53. LEACH • LEACH adopts a hierarchical approach to organize the network into a set of clusters • Cluster is managed by a selected cluster head • The cluster head carry out multiple tasks – Periodic collection of data from the members of the cluster – Aggregates – Transmit the aggregated data to the base station(Single hop) – Create a TDMA-based schedule for each node of the cluster 53
- 55. Basic operations of LEACH • Two distinct phases – The first phase • Setup phase, consists of two steps – Cluster-head selection – Cluster formation – The second phase • Steady-state phase focuses on – Data collection, aggregation, and delivery to the base station – The duration of the setup is assumed to be relatively shorter than the steady-state phase to minimize the protocol overhead. 55
- 56. LEACH phases 56
- 57. Cluster-head selection process • The cluster-head selection process ensures that this role rotates among sensor nodes • Node, n, generates a random number, v, between 0 and 1 • Compares v to the cluster-head selection threshold T(n) • The node becomes a cluster head if v is less than T(n) 57
- 58. Cluster-head selection process • The cluster-head selection threshold is designed to ensure – High probability that a predetermined fraction of nodes, P, is elected cluster heads at each round – Nodes which served in the last 1/P rounds are not selected in the current round. 58
- 59. Cluster-head selection process Threshold T(n) of a competing node n expressed as follows: G : Represents the set of nodes that have not been selected to become cluster heads in the last 1/P rounds r : Denotes the current round P : Represents the cluster-head probability (predefined). It is clear that if a node has served as a cluster head in the last 1/P rounds, it will not be elected in this round. 59
- 60. LEACH CH broadcast is selection CH creates and distributes the TDMA schedule for each member of the cluster CH selects a CDMA code and distributed to all members The code is selected carefully so as to reduce inter-cluster interference. The completion of the setup phase signals the beginning of the steady- state phase. Steady-state phase Nodes collect information and use their allocated slots to transmit to the cluster head 60