Routing Hole Handling Techniques for Wireless Sensor Networks: A Review
محورهای موضوعی : Computer Networks and Distributed Systems
Swathi B H
1
(Vidyavardhaka College of Engineering)
Gururaj H L
2
(Vidyavardhaka College of Engineering)
کلید واژه: Geographic routing, Wireless Sensor Networks (WSNs), routing holes, sensor nodes, greedy forwarding,
چکیده مقاله :
A Wireless Sensor Network consists of several tiny devices which have the capability to sense and compute the environmental phenomenon. These sensor nodes are deployed in remote areas without any physical protections. A Wireless Sensor Network can have various types of anomalies due to some random deployment of nodes, obstruction and physical destructions. These anomalies can diminish the sensing and communication functionalities of the network. Many kinds of holes can be formed in a sensor network that creates geographically correlated areas. These holes are also responsible for creating communication voids. These voids do not let the packets to reach the destination and minimises the expected network performance. Hence it ought to be resolved. In this paper we presented different kinds of holes that infect the sensor network, their characteristics and the effects on successful communication within the sensor network .Later we presented a detailed review on different routing hole handing techniques available in literature ,their possible strengths and short comes. At last we also presented a qualitative comparison of these routing hole handing techniques.
1. Torrieri, D., S. Talarico, and M.C. Valenti, Performance comparisons of geographic routing protocols in mobile ad hoc networks. IEEE Transactions on Communications, 2015. 63(11): p. 4276-4286.
2. Mostefaoui, A., M. Melkemi, and A. Boukerche, Localized routing approach to bypass holes in wireless sensor networks. IEEE transactions on computers, 2013. 63(12): p. 3053-3065.
3. Petrioli, C., et al., ALBA-R: Load-balancing geographic routing around connectivity holes in wireless sensor networks. IEEE Transactions on Parallel and Distributed Systems, 2013. 25(3): p. 529-539.
4. Cadger, F., et al., A survey of geographical routing in wireless ad-hoc networks. IEEE Communications Surveys & Tutorials, 2012. 15(2): p. 621-653.
5. Lima, M.M., H.A. Oliveira, and R.W. Pazzi. A novel RSSI-based algorithm for detect and bypass routing holes in wireless sensor networks. in 2018 IEEE Symposium on Computers and Communications (ISCC). 2018. IEEE.
6. Anchugam, C. and K. Thangadurai, Detection of Black Hole Attack in Mobile Ad-hoc Networks using Ant Colony Optimization-simulation Analysis. Indian Journal of Science and Technology, 2015. 8(13): p. 1-10.
7. Das, S. and M.K. DebBarma, Hole Detection in Wireless Sensor Network: A Review, in Recent Findings in Intelligent Computing Techniques. 2018, Springer. p. 87-96.
8. Zhou, F., et al., Bypassing holes in sensor networks: Load-balance vs. latency. Ad Hoc Networks, 2017. 61: p. 16-32.
9. Vu, H., et al. An efficient geographic algorithm for routing in the proximity of a large hole in wireless sensor networks. in Proceedings of the Seventh Symposium on Information and Communication Technology. 2016.
10. Ammari, H.M., Investigating the energy sink-hole problem in connected $ k $-covered wireless sensor networks. IEEE Transactions on Computers, 2013. 63(11): p. 2729-2742.
11. de Oliveira, H.A., et al., An enhanced location-free Greedy Forward algorithm with hole bypass capability in wireless sensor networks. Journal of Parallel and Distributed Computing, 2015. 77: p. 1-10.
12. Lima, M.M., et al., Geographic routing and hole bypass using long range sinks for wireless sensor networks. Ad Hoc Networks, 2017. 67: p. 1-10.
13. Liu, J., et al., Characterizing data deliverability of greedy routing in wireless sensor networks. IEEE Transactions on Mobile Computing, 2017. 17(3): p. 543-559.
14. Ren, J., et al., Lifetime and energy hole evolution analysis in data-gathering wireless sensor networks. IEEE transactions on industrial informatics, 2015. 12(2): p. 788-800.
15. Latif, K., et al., On energy hole and coverage hole avoidance in underwater wireless sensor networks. IEEE Sensors Journal, 2016. 16(11): p. 4431-4442.
16. Chen, D. and P.K. Varshney, A survey of void handling techniques for geographic routing in wireless networks. IEEE Communications Surveys & Tutorials, 2007. 9(1): p. 50-67.
17. Huang, H., et al., Coordinate-assisted routing approach to bypass routing holes in wireless sensor networks. IEEE Communications Magazine, 2017. 55(7): p. 180-185.
18. Fang, Q., J. Gao, and L.J. Guibas. Locating and bypassing routing holes in sensor networks. in IEEE INFOCOM 2004. 2004. IEEE.
19. Karp, B. and H.-T. Kung. GPSR: Greedy perimeter stateless routing for wireless networks. in Proceedings of the 6th annual international conference on Mobile computing and networking. 2000.
20. Rehman, A.-u., S.U. Rehman, and H. Raheem, Sinkhole Attacks in Wireless Sensor Networks: A Survey. Wireless Personal Communications, 2019. 106(4): p. 2291-2313.
21. Ramazani, S., et al., Topological and combinatorial coverage hole detection in coordinate-free wireless sensor networks. International Journal of Sensor Networks, 2016. 21(1): p. 40-52.
22. Wood, A.D., J.A. Stankovic, and S.H. Son. JAM: A jammed-area mapping service for sensor networks. in RTSS 2003. 24th IEEE Real-Time Systems Symposium, 2003. 2003. IEEE.
23. Amgoth, T. and P.K. Jana, Coverage hole detection and restoration algorithm for wireless sensor networks. Peer-to-Peer Networking and Applications, 2017. 10(1): p. 66-78.
24. Kaur, G., V. Jain, and Y. Chaba. Detection and Prevention of Blackhole Attacks in Wireless Sensor Networks. in International Conference on Intelligent, Secure, and Dependable Systems in Distributed and Cloud Environments. 2017. Springer.
25. Benzerbadj, A., et al., Cross-Layer Greedy position-based routing for multihop wireless sensor networks in a real environment. Ad Hoc Networks, 2018. 71: p. 135-146.
26. Huang, H., et al., Energy-aware dual-path geographic routing to bypass routing holes in wireless sensor networks. IEEE Transactions on Mobile Computing, 2017. 17(6): p. 1339-1352.
27. Yu, Y., R. Govindan, and D. Estrin, Geographical and energy aware routing: A recursive data dissemination protocol for wireless sensor networks. 2001.
28. Kranakis, E., H. Singh, and J. Urrutia. Compass routing on geometric networks. in in Proc. 11 th Canadian Conference on Computational Geometry. 1999. Citeseer.
29. Bose, P., et al., Routing with guaranteed delivery in ad hoc wireless networks. Wireless networks, 2001. 7(6): p. 609-616.
30. Kuhn, F., et al. Geometric ad-hoc routing: Of theory and practice. in Proceedings of the twenty-second annual symposium on Principles of distributed computing. 2003.
31. De Couto, D.S. and R. Morris, Location proxies and intermediate node forwarding for practical geographic forwarding. 2001, MIT Laboratory for Computer Science Technical Report MIT-LCS-TR-824.
32. Sadeghzadeh-Nokhodberiz, N. and J. Poshtan, Distributed interacting multiple filters for fault diagnosis of navigation sensors in a robotic system. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2016. 47(7): p. 1383-1393.
33. Phan, H.-P., et al., The piezoresistive effect of SiC for MEMS sensors at high temperatures: a review. Journal of Microelectromechanical systems, 2015. 24(6): p. 1663-1677.
34. Salkuyeh, M.A. and B. Abolhassani, An adaptive multipath geographic routing for video transmission in urban VANETs. IEEE Transactions on Intelligent Transportation Systems, 2016. 17(10): p. 2822-2831.
35. Zhu, L., et al., Geographic routing in multilevel scenarios of vehicular ad hoc networks. IEEE Transactions on vehicular technology, 2015. 65(9): p. 7740-7753.
36. Awang, A., et al., Routing in vehicular ad-hoc networks: A survey on single-and cross-layer design techniques, and perspectives. IEEE Access, 2017. 5: p. 9497-9517.
37. Jin, X., et al., TIGHT: A geographic routing protocol for cognitive radio mobile ad hoc networks. IEEE Transactions on Wireless Communications, 2014. 13(8): p. 4670-4681.
38. Jung, W.-S., J. Yim, and Y.-B. Ko, QGeo: Q-learning-based geographic ad hoc routing protocol for unmanned robotic networks. IEEE Communications Letters, 2017. 21(10): p. 2258-2261.
39. Zhou, H., et al., Localized and precise boundary detection in 3-D wireless sensor networks. IEEE/ACM Transactions on Networking, 2014. 23(6): p. 1742-1754.
40. Zhang, D. and E. Dong, A virtual coordinate-based bypassing void routing for wireless sensor networks. IEEE sensors journal, 2015. 15(7): p. 3853-3862.
41. Coutinho, R.W., et al., Geographic and opportunistic routing for underwater sensor networks. IEEE Transactions on Computers, 2015. 65(2): p. 548-561.
42. Hasan, M.Z., F. Al-Turjman, and H. Al-Rizzo, Analysis of cross-layer design of quality-of-service forward geographic wireless sensor network routing strategies in green internet of things. IEEE Access, 2018. 6: p. 20371-20389.
43. Zhu, C., et al., Sleep scheduling for geographic routing in duty-cycled mobile sensor networks. IEEE Transactions on Industrial Electronics, 2014. 61(11): p. 6346-6355.
44. Zhang, D. and E. Dong, An efficient bypassing void routing protocol based on virtual coordinate for WSNs. IEEE Communications Letters, 2015. 19(4): p. 653-656.
45. Shu, L., et al., Geographic routing in duty-cycled industrial wireless sensor networks with radio irregularity. IEEE Access, 2016. 4: p. 9043-9052.
46. Huang, H., et al., Three-dimensional geographic routing in wireless mobile ad hoc and sensor networks. IEEE Network, 2016. 30(2): p. 82-90.
47. Gupta, H.P., et al., Geographic routing in clustered wireless sensor networks among obstacles. IEEE sensors Journal, 2014. 15(5): p. 2984-2992.
48. Popescu, A.M., N. Salman, and A.H. Kemp, Energy efficient geographic routing robust against location errors. IEEE Sensors Journal, 2014. 14(6): p. 1944-1951.
49. Cheng, L., et al., QoS aware geographic opportunistic routing in wireless sensor networks. IEEE Transactions on Parallel and Distributed Systems, 2013. 25(7): p. 1864-1875.
50. Ghoreyshi, S.M., A. Shahrabi, and T. Boutaleb, Void-handling techniques for routing protocols in underwater sensor networks: Survey and challenges. IEEE Communications Surveys & Tutorials, 2017. 19(2): p. 800-827.
51. Khalifa, B., et al., Coverage hole repair in WSNs using cascaded neighbor intervention. IEEE Sensors Journal, 2017. 17(21): p. 7209-7216.
52. Yan, F., et al., Homology-based distributed coverage hole detection in wireless sensor networks. IEEE/ACM transactions on networking, 2014. 23(6): p. 1705-1718.
Routing Hole Handling Techniques for Wireless Sensor Networks: A Review
Abstract: A Wireless Sensor Network consists of several tiny devices which have the capability to sense and compute the environmental phenomenon. These sensor nodes are deployed in remote areas without any physical protections. A Wireless Sensor Network can have various types of anomalies due to some random deployment of nodes, obstruction and physical destructions. These anomalies can diminish the sensing and communication functionalities of the network. Many kinds of holes can be formed in a sensor network that creates geographically correlated areas. These holes are also responsible for creating communication voids. These voids do not let the packets to reach the destination and minimises the expected network performance. Hence it ought to be resolved. In this paper we presented different kinds of holes that infect the sensor network, their characteristics and the effects on successful communication within the sensor network .Later we presented a detailed review on different routing hole handing techniques available in literature ,their possible strengths and short comes. At last we also presented a qualitative comparison of these routing hole handing techniques.
Key words: Wireless Sensor Networks (WSNs), Geographic routing, greedy forwarding, sensor nodes, routing holes
1. Introduction
Now a days, Micro Electrical Mechanical System (MEMS) [32][33] and Wireless Communication Technology is gaining global attention in research area of Wireless Sensor Networks (WSNs). A sensor network consists of several tiny and economical nodes [10]. These sensor nodes are organized in such a way that they can sense some phenomenon, evaluate and store the sensed information and can communicate with the other nodes .These sensor networks can be used in several applications like military applications, intrusion detection, environmental monitoring, hazard detection, heath related applications, meteorology, agriculture, industry applications, infrastructure monitoring and many other [14]. Even though they have many useful applications it suffers from some constraints like low processing speed, limited bandwidth, less memory and limited battery. These constraints present various operational, management and design issues [7][25].
The main task of a sensor node in WSNs is to sense the particular phenomenon and carryout communication among the sensor nodes. To achieve the expected communication between these nodes, the nodes should be deployed in such a way that it should cover 100% of the target field. Various types of anomalies can occur during the deployment of nodes and above mentioned constraints can cause a sensor node to fail. The area with a failed node causes a hole in the network which is a major issue that ought to be solved. Because, occurrence of the holes causes the communication delay and potentially degrades the quality of the service in WSNs. There are various reasons behind the creation of holes in the sensor networks .Holes may occur during the deployment stage of the network, during continues operation or by some environmental phenomenon [8].These routing holes leads to performance degradation of network and the communication between nodes become weak. Hence routing of packet in such scenario is the great challenge. Therefore identifying and recovering a hole is very important [3][12].
Geographic routing is an efficient and widely accepted routing approach to resolve the routing hole problem[17][40]. In geographic routing, packets are delivered to the intended destination using the location or the geographic position of the destination rather than logical or identity of the destination. There is no need to maintain entire routes information from source node to destination node. Nodes need not to store routing tables[42][43]. Because, it uses only one-hop geographic information of neighbour node. It also supports geocasting services. Because of its stateless and localized features the geographic routing approach is considered as scalable and simple. Performances of various geographic routing protocols are described in literature [1]. In 1980's geographic routing approach was basically proposed for packet radio networks which is also known as location based routing, position based routing or directional routing approach. As the applications of self configuring localization mechanism and Global Positioning System (GPS) are increasing, it is gaining more attention and providing efficient solution for WSNs, Underwater sensor[41][50], Vehicular AdHoc Networks (VANETs) [34][35][36], Mobile AdHoc Networks (MANETs)[37][38][46] and Wireless Mesh Networks (WMNs) [12][16][39] .
The remaining part of the paper is structured as follows: In section 2 we have discussed the categories of hole in wireless sensor networks and the basic principles of geographic routing approaches. Section 3 provides an elaborated description about the hole detection and recovery approaches especially on routing holes with respect to various literature works. Summarization of different techniques is done in section 4. And finally section 6 concludes.
2. Categories of holes in WSNs
This section presents different types of holes that a WSNs can have and the basic principles of hole avoiding strategies.
2.1 Holes for WSNs
A wireless sensor network can have different types of holes including coverage hole, jamming hole, routing hole, and sink hole/black hole/warm hole. Coverage hole can occur due to the failure of node, due to power exhaustion, fault in network topology and random deployment of sensor nodes that create voids. This leads to coverage hole and routing hole within the sensor network [15][21][23][51][52]. Sometimes the nodes in the network may not be able to communicate with each other according to the designed goal to reach their objectives. This happens if a node cannot sense or relay the sensed information to the destination. These types of holes are called as routing hole [19]. Routing hole can also occur if a fresh sensor node is replaced in the place of a faulty sensor node. Because it may changes the routing path in the network. A jammer is used to interrupt the communication between the nodes. The jammer can block the radio frequencies by using high-frequency signals. These are called as jamming holes. Jamming holes may also occur when the malicious node tries to jam the communication between the nodes [22]. Other holes like sink hole/black hole can occur due to the denial of service attack where the malicious node may mislead the packet on routing path [6] [20][24] .
2.2 Basic principles of Geographic routing approach
Geographic routing approach has been proposed and widely accepted method to overcome from the routing issues since it is scalable and uses localized information[44][45][47][48]. To overcome from the issues of routing holes , geographic routing uses two well known fault tolerance mechanisms namely, greedy forwarding and face routing[4][9][26][27][28].
Fig.1 Example for greedy forwarding
The greedy forwarding is the simple and basic approach. Here at each hop, positive progress is carried out to route the packet to the neighbour node which is located closer to the intended destination node. When a current node is not able to identify a neighbouring node located closer to the destination node than itself it should drop the packet which is being forwarded. That is, a source node fails to identify a next-hop using positive progress towards the destination. This unenviable phenomenon is called as local minimum or local maximum phenomenon. It is often referred as communication void. [11][13][16][19][24][25].
Fig.1 depicts the example of greedy forwarding approach. The blue circles in the diagram represent range of transmission of a node. The black small circles represent the nodes and a green line indicates the packet routing path. The node S indicates the source and it follows the principle of greedy forwarding and sends the packet to the node which is nearer to the destination D than itself. A node which receives the packet will carry out the same procedure. The Fig.1 is a just an idealized scenario but it is somewhat impossible in real networks because of the dynamic nature of the network. Hence it is not possible to identify a nearest node of a destination. Hence the packet might be dropped even though there is a path from the source node to destination node. Hence it leads the node into local maximum phenomenon. Even though greedy forwarding is well understandable, easy to implement and being more efficient it suffers from this major drawback [4].
The face routing approach is basically inherited from Compass routing II [28]. This approach is come up as a healing mechanism from local maximum phenomenon and as a well known routing approach. By combining the possible strengths of both greedy forwarding method and face routing method the disadvantages of greedy forwarding can be minimized. Face routing uses a process called Planarization to construct planer sub graphs. A planer graph is one which consists of no crossing links and contains sequence of polygonal areas separated by faces or edges. These planner graphs can be obtained by using Gabriel Graph (GG) or Relative Neighbourhood Graph (RNG) [4][16].
Fig.2 Example for face routing approach
Face routing uses right hand or left hand rule to traverse these planar sub-graphs. A basic operation of face routing is shown in Fig.2. The node S represents the sender and the node D represents the destination. The thin black lines represent the faces which are obtained by planarization algorithm. The small circle indicates a point which intersects the edges of the graph [19]. The traversed faces on the routing path from S to D are numbered sequentially F1 to F5. The routing starts from source node S and it uses right hand rule to traverse face F1 with the point closest to the D which intersects the S-D line being the point where the next face is traversed. F2 is then traversed before the point closest to D that intersects the line S-D is discovered .When it completes the traversal of entire face, the algorithm goes to the any one of the intersection which is closest to the destination and continues to do so until eventually the D is reached. The major advantage with face routing is guarantee delivery of packets. However its possible inefficiency is the major drawback [4].
3. Routing hole avoidance techniques
This section provides a overview of the various types of techniques which are used to bypass routing hole in wireless sensor networks using geographic routing approach.
3.1 Greedy Perimeter Stateless Routing
Greedy Perimeter Stateless Routing (GPSR) [19] was mainly proposed to increase the scalability under the scenarios of increasing rate of nodes in the network and increasing rate of mobility. It uses greedy forwarding to send the packets from source to destination as illustrated in section 3. Packet forwarding decisions can be made by using the location of router and location of destination. As a result, an intermediate packet sending node can choose packet’s next hop by making locally optimal and greedy choice. Means, if a node has a radio neighbour location information, it can choose that node as a next hop by using local optimal choice if the neighbour node geographically nearer to the destination.
The greedy forwarding only depends on the information about the immediate neighbour of the packet forwarder node. It doesn't require state information. Hence GPSR is stateless. It is dependent on the density of the nodes but not on the total number of destination nodes in WSNs. The disadvantage of greedy forwarding is that it suffers from local minimum phenomenon which is shown in Fig. 3.
Fig.3 Failure of greedy forwarding, nodes a and c are geographically farther from destination D, node b is a local minimum in its proximity to node D.
Here, the node b has a packet to send to the destination node D. Node b is nearer to destination node D than its neighbour nodes a and c. The arc about the node indicates the range of the node. The distance between D and b is equal to the radius of D. Even though D has two paths (b->c->e->D) and (b->a->f->D), the node b doesn't choose a or c based on greedy forwarding. Node b is a local minimum to D in this situation and some other strategies must be chosen to route the packets.
Fig.4 Node b's void with respect to D.
The intersection between the circular range of node b and the radius of D is empty without any neighbouring nodes. It is illustrated in Fig.4. From the node b's perspective the shaded area without any nodes is a void.
Perimeter forwarding approach will be applied in such cases when greedy forwarding method fails to forward the current sending packet. It is also called as face routing approach [8]. Perimeter forwarding can be applied for planarized graph where as greedy forwarding can be applied on full network graph. Planar graph uses both interior faces and exterior faces to transmit the packet using right hand rule. Fig.5 shows an example for perimeter forwarding. In the figure, each traversed face is penetrated by aD. The first two and last faces indicates interior faces, third is the exterior face.
Fig.5 Example for Perimeter forwarding
In static network topology, the no-crossing heuristic identifies the maximum number of routes but it is difficult to identify a reachable route to node. Hence planer graph can be used to remove crossing links. But the graph must not be separated by removal of crossing edges. The Relative Neighbourhood Graph (RNG) and Gabriel Graph (GG) are the two well known and commonly used planar graphs which shown in Fig.6 and Fig.7 respectively.
Fig.6 RNG graph Fig.7 Gabriel Graph
RNG is defined as “An edge Em,n exists between vertices m and n if the distance between the vertex is less than or equal to the distance between every other vertex p and whichever of m and n is farther from vertex p". In equation form.
∀ p≠m,n:d(m,n) ≤ max[d(m,p),d(n,p)] (1)
A RNG graph can be constructed by removing the crossing links as illustrated in Algorithm 1.
Algorithm : 1 Relative Neighbourhood Graph |
1. for all n € N do 2. for all p € N do 3. if p = = n then 4. continue 5. else if d(m,p) > max[d(m,p),d(n,p)] then 6. remove edge(m,n) 7. break 8. end if 9. end for 10. end for
|
GG is defined as- "An edge Em,n exists between vertices m and n if there is no vertex p is there within the circular region whose diameter is mn". In equation form
∀ p≠m,n:d2(m,n) ≤ [d2(m,p),d2(n,p)] (2)
GG graph can be constructed by removing the crossing links as in the Algorithm-2 .
Algorithm : 2 Gabriel Graph |
1. c=centrepoint of edge mn 2. for all n € N do 3. for all p € N do 4. if p = = n then 5. continue 6. else if d(c,p) < d(m,c) then 7. remove edge(m,n) 8. break 9. end if 10. end for 11. end for
|
The well known Right hand rule is depicted in Fig.8. It states that when a packet sent from c to node a, it traverse the next edge (a,b) sequentially in counter clock wise about a from edge(a,c).The right hand rule will traverse the interior part of a closed polygon region or a face in clock wise order. In Fig.8 the nodes a,b,c creates a triangle in the order(c->a->b->c).
Fig.8 Right hand rule
The disadvantage of GPSR is that, even though it bypasses holes in the network which are caused by local minimum phenomenon, it has to maintain planar graph on each node which leads to overhead on nodes. Because, these information is used by only those nodes which are affected by local minimum phenomenon.
3.2 Compass Routing II
Karnakis et at.[28] proposed the Compass Routing II algorithm. It guarantees the packet delivery to destination node even if the node is in a local minimum problem in greedy forwarding. It uses face routing with least deviation angle of the link between the current node and the destination node when it is trying to forward a packet to the next hop. As Compass Routing II [28] , FACE-I and FACE-II algorithms are proposed [29] that guarantees the packet delivery. It also uses Gabriel Graph(GG) to construct planer sub graph and then uses right hand rule to traversal the faces. FACE-I is modified in FACE-II in which the traversal of face will follow the next edge wherever that link crosses the line from source node to destination node. The routing hole issue can be solved by always using the face routing.
3.3 Greedy Other Adaptive Face Routing
Greedy Other Adaptive Face Routing (GOAFR, GOAFR+) [30] was introduced as the extension of Compass Routing II. It has come up with the fall back scheme to come back to the perimeter forwarding mode without exploring the entire face boundary if it is not necessary. Same as [19],[28] it constructs the planar sub graph using Gabriel Graph. It begins with greedy forwarding mode and whenever it encounters a local minimum phenomenon, it will switch to face routing mode. It uses two counters. First is to keep track of the number of attended nodes during face routing that are closer to the destination node. Another one is to keep track of the number of nodes which are far from the destination node. These two counters are used to decide whether it has to continue in the face routing or fall back to greedy mode. Algorithm 3 illustrates the steps involved in GOAFR+.
Algorithm : 3 Greedy Other Adaptive Face Routing |
Require : Choose Algorithm parameters to ρ ,σ before the beginning of algorithm and remains same during execution and set the conditions 1≤ ρ0< ρ and 0< σ and also two counters x, y
1. Initiate from source node S and Z as a circle with centre D where rD = ρ0|sD| 2. Continue greedy forwarding until it reaches D or occurrence of local minimum // Terminate if true. 3. Begin face routing and explore the boundary of F1 face has a line aiD //where ai is the current local minimum 4. If visited node is nearer to D then repeat step 2.Otherwise report node S as graph has disconnected. 5. Use x to indicate the node nearer to D than ai and y to indicate node which is far from D than ai and has to take some special actions if following cases meet Case 1: If Z hits for the first time, go back and explore the F1's boundary in the reverse direction Case 2: If Z hits for the second time and there is no visited node nearer to D than ai increase radius of Z (rz:= ρrz) repeat step 2 if it began from ai. Otherwise repeat with step 2. Case 3: If x > σ y, then it says that the number of visited nodes which are nearer to D are more than the nodes which are not. Repeat step 2.
|
Consider the Fig.9 which illustrate GOAFR+ algorithm with parameter σ, x, y where x and y are the counters and σ is the condition for fall back which is set to be 1/4 ≤ σ ≤1/2.
Fig.9 Example for GOAFR+
The node S begins to forward packet in greedy mode. The packet reaches node a, when it doesn't get any neighbour node nearer to the destination and it will move to face routing. It begins exploring the boundaries of face F. At node b, the algorithm reaches the boundary of the circle Z and come back to exploring the face F in opposite direction. Counters x, y keep on updating after each step. At node c, the number of nodes visited so far which are nearer to the destination D is two. Hence the counter value a=2. The number of nodes visited that are far away to the node D is four. Hence counter value b=4. So the fall back condition is true, i.e., a> σ b. Therefore, the algorithm fall back to greedy forwarding mode and continues the process until it reaches the destination.
3.4 Geographic and Energy Aware Routing
Geographic and Energy Aware Routing (GEAR) [27] was proposed to forward a packet to the intended destination. It works in two steps of operation. In first step, the packet will be forwarded to the destination by making use of energy aware next-hop neighbour selection. In second step, the packets are distributed inside the region using recursive geographical or restricted flooding. In first step, two types of costs are maintained by each node. They are estimated cost and a learned cost for reach ability of the destination node via its neighbour nodes. The one-hop neighbour information is used to obtain the learned cost of a region and by adding it to the particular selected neighbour cost a node can calculates its learned cost.
The estimated cost is used as default cost if there is no learned cost is available. The estimated cost depends on the distance to the destination node and the residual energy of a node. The presence of hole in the network will reflects the variation of the routing path on the new learned cost which encounters the routing hole. If there is no hole appears, the learned cost is equal to the estimated cost. The recursive geographical forwarding is used in case of high density sensor network instead of restricted flooding. The advantage of GEAR is that it works well in the small sensor area network.
3.5 Intermediate Node Forwarding
A probabilistic approach called Intermediate Node Forwarding was introduced by De. Couto et al.[31] to forward packets around routing hole by assuming unequal radio ranges. When packet drop occurs due to local minimum the source node will get the feedback using Negative Acknowledgement (NAKs).
When the sender gets the information about the packet drop, it will select any location randomly from a disk that is one quarter distance between the sender node and the destination node, which is centred at the middle point of the distance between the sender and the destination. If the packet is dropped by the chosen node to be doubled along with another randomly selected intermediate geographic location. The disadvantage of this approach is, the protocol overhead may occur due to NAKs.
3.6 Boundhole
Qing Fang et al.[18] proposed an approach to define the nodes as stuck nodes that are caused by the local minimum phenomenon. A local rule called TENT rule is developed for each node of the network to verify the presence of the stuck node. If the node is identified as stuck node, then it uses BOUNDHOLE-a distributed algorithm to identify the holes. Similar to the planar graph of the perimeter routing, the boundary information which is stored locally is used to get rid from the local minimum phenomenon. It computes and stores information of holes only at the region of routing hole. In order to identify the stuck node the TENT rule can be implemented.
Fig.10 The TENT rule to identify a void
As depicted in Fig.10 for each node p, we order its entire one-hop neighbour in counter clockwise. The adjacent nodes m and n have a perpendicular bisector mp and np which intersects the point X. If the point X is in the range of p there is no stuck node. Because m and n are adjacent and there is no node within the black region. Hence node p cannot be a stuck node with the cone pm and pn. Conversely, if X is in the outside of the communication range there will be a stuck node at p. So TENT rule is necessary to identify the void node.
Fig.11 BOUNHOLE approach
After the identification of the of the stuck node, the BOUNDHOLE approach is used by the node p. The Fig. 11 shows the basic idea. The nodes s, p, t1, t2 bounds the hole. This approach begins from stuck node p and sweep in the direction of stuck node. It sends packet to a neighbour node t1 in clockwise direction. The t2 receives the packet from t1 and t2 passes it to s. In general this process repeats until the packet reaches back to p by marking the boundary of the hole. Hence similar to the perimeter routing it creates a conduit for the stuck packets that are routed in the void area.
Fig.12 (i) Fig.12 (ii)
Fig.12 An example to illustrate how the approach utilizes hole surrounded path.Fig-12(i) The destination node d leis outside the hole Fig-12(ii) The destination node d lies inside the hole.
Consider the Fig.12 (i) where node p is a void node where the packet is stuck when it is in the greedy forwarding mode. The boundary of the hole may contain both void and non void nodes. Hence according to the BOUNDHOLE approach the node p must be on the boundary. The void node forwards the packet on this boundary. The packet is forwarded greedily again if the routed packet is received by the closest node of the destination d than p. The example is shown the Fig.12 (i). Here the destination d is in the outside region the hole and a node m that is nearer to the destination d must exit than p. Suppose if we join the line pd, it crosses the edge mn which is a hole boundary. The packet will always reach the destination because both node m and n are closer to node d and p. If the node d is inside the hole region then restricted flooding can be applied and it is depicted in Fig12(ii). The disadvantage of BOUNDHOLE is it impose higher load on nodes which are the boundaries of hole.
3.7 Curved Stick
Ahmed Mostefaoui et.al [2] proposed a novel approach to overcome the disadvantages of BOUNDHOLE called Curved Stick (CS). The basic idea of Curved Stick (CS) method is to characterize the nodes which are responsible for a communication intersection situation which is depicted Fig.13. A communication intersection says that, it is a situation when a edge E1,2 is on the boundary of the nodes N1 and N2 that crosses by an edge E3,4 where N1 and N2 are not visible to N3 . Hence the basic idea behind this is to overcome from the false boundary identification problem of BOUNDHOLE problem. It uses Curved Stick to select the next hop rather than sweeping line.
Fig.13: Communication intersection Problem
The CS routing algorithm works mainly in three phases a) Engaging phase b) CS boundary traversal and c) Termination phase
The engaging phase is similar to the greedy forwarding approach. The packets between the nodes use the geographical information to reach the destination. If the packet cannot be delivered to the destination then it encounters local minimum phenomenon. As shown in Fig-14 Ninit is the initiator node where the message got stuck. Then it implements the CS boundary traversal phase.
In this phase, every attended node collects the initiator’s location information from the earlier attended hop node. It uses this information as an input. Initially, the visited node figure outs its distance to the destination and from the initiator node to the destination. The packet will considered as local minimum problem free if its distance is closer to the destination than the initiator node. If it is out of local minimum problem then it can use greedy forwarding approach. Otherwise CS rule can be applied.
Fig.14 Curved Stick Rule
The termination phase can takes place only in two cases. In the first case, the destination must have received the packet. In the second case, the packet uses the CS rule to travel on the entire boundary and gets back to Ninit. Here, the curved stick is swept by the initiator. If the node is hit, it is considered as the next hop node and the CS rule will continue. If not the Ninit cannot implement CS rule and it says that there is no route from the source to destination and the initiator indicates the source node that the packet cannot be sent. They presented a hop count reduction method which can be applied when there are multiple sweeping direction. However CS rule can bypass hole but still it suffers from local minimum problem. The disadvantage of this CS rule is that it can be applicable only for two- dimensional network and it still suffers from local minimum problem. Algorithm 4 illustrate the Curve Stick routing.
Algorithm : 4 CS Routing Algorithm |
1. if (Msg.Init.ID = null) then //where Msg is the message from previous hop 2. use greedy forwarding 3. select the next hop based on greedy forwarding 4. if(stuck node occurs) then 5. set Msg.Init.ID to MyID 6. select next hop using CS rule 7. end if 8. else 9. if Msg.Init.ID ≠ MyID then 10. calulate:d(Ncur,Nd) and d(Ninit,Nd) // where Ncur is the current node and Nd is the destination node 11. if d(Ncur,Nd) < d(Ninit,Nd) then //Free of local minimum phenomenon 12. set Msg.Init.ID to null 13. select next hop using greedy forwarding 14. else 15. select next hop using CS rule 16. end if 17. else // message will be traversed entire boundary and returned to the initiator sweep the //curved stick from starting point 18. if a node hits then 19. select a next hop using CS rule 20. else 21. notify the source node that the message could not be reached 22. end if 23. end if 24. end if
|
3.8 Long Range Sink
Moyses M. Lima et.al [5] have proposed an algorithm Long Range Sink(LRS) that considers the greater communication capability of sink node and the Received Signal Strength Indicator(RSSI) by constructing the packet forwarding path from source to destination. It uses the principle of forwarding the packet to the neighbour nodes that obtained the query with higher signal strength.
Initially it uses greedy forwarding technique along with LRS. When packets reaches to a hole region it will switch to hole-election mode. It performs data aggregation using hop-election mechanism to bypass the routing hole.
At the beginning sink sends the queries to every other nodes in a single hop communication. The RSSI technique is used to calculate the distance between sink and each node. All the information about the distance are exchanged between the neighbour nodes. The data of regular node will be aggregated to the list of aggregated data if it already contains any data that is related to the query of the sink node.
One hop neighbour of the sink node broadcasts the additional information to its immediate neighbour node. This additional information is used to identify the routing path from sink to the current node via the neighbour node that broadcasts the information. An initialization timer is used to keep track the performance of the data aggregation.
The nodes with lesser RSSI are omitted to avoid the participation of the nodes in a alternative hop-election. Whenever the initialization timer expires, a replay message is combined with both the node's distance information and aggregated data. Later, the node forwards this information to the neighbour node which is closer to the sink node. The message received by the neighbour of the sink aggregates the data with the list of aggregated data and forwards when the timer expires. The above process performs repeatedly until the packet reaches the sink node. When sink dispatches a query to the entire network it creates a eligible nodes table using which each node tries to identify whether it is in a hole region or not during the process of packet forwarding. If it is a hole node it chooses a new neighbour node towards the sink node using the extra information which is saved during the evaluation of distance. Then the hole node aggregates the packets from the hole region and sends packet via multiple hops toward sink.
3.9 Energy -aware Dual-path Geographic Routing
In this paper a new routing protocol Energy-aware Dual-path Geographic Routing (EDGR) is proposed by Haojun Huang et.al [26] to overcome from the more energy conservation and load balance issues of existing approaches. EDGR approach uses greedy mode to route the packet in two various routing paths instead of bypass mode. It creates dual routing paths using two node-disjoint anchor lists. Since it makes the packet to travel through double sides of the holes, it shortens the path length for routing and balances the load. EDGR selects the node with higher residual energy as packet forwarders by using two node-disjoint anchor list. Hence the data packets are routed along two paths of the hole.
The EDGR works in two operation phases: obtaining anchor list and data dissemination phase. In the first phase two anchor lists are obtained by using adaptive approach based on the distance of nodes. In the second phase, the routing decisions are taken care by using geographic information, energy consumption characteristics and residual energy. Algorithm 5 illustrate the building of anchor list.
EDGR uses three types of packets namely: burst packet, beacon packet and data packets. Using burst packet, anchor lists can be identified. The beacon packets are used to exchange the residual energy and location information.
Fig-15 Burst packet format
Algorithm : 5 Building anchor list |
Require: Source s, destination d
1. list(s,d)=[flag,Ø] 2. s starts exchange of beacon packets 3. s attaches list(s,d) to a burst packet and sends to d 4. if ∀ Vi in the bypass mode receive the packet then // Vi is any forwarder node 5. UpdateList(Vi,list(s,d)) 6. end if 7. if s gets a feedback packet from d then 8. update list(s,d)=[flag,a1....aj....,am] 9. forward a burst packet via a1,a2,a3....,am 10. end if 11. if ∀ Vi = = ak€{a1,....,am} obtains the packet then 12. forward the packet to ak+1 13. else 14. if ∀ Vi is in the bypass mode then UpdateList(Vi,list(s,d)) 15. end if 16. function(UpdateList(Vi,list(s,d)) 17. if Vi € list(s,d) then 18. remove each node after Vi in list(s,d) 19. end if 20. if Vi = = Vfj or Vi= =Vlj or Vi-1, Vi+1 are in bypass mode then 21. compute candidate Vi 22. add Vi to list(s,d) 23. send list(s,d) to Vi+1 following flag 24. end if 25. if Vi == d then 26. send list(s,d) to s 27. end if 28. end function
|
The Fig.15 depicts the burst packet format. It has source and destination locations and a list of anchor nodes. This list consists of a sequence of anchor nodes. A flag field represents whether the data packet bypass the routing hole or not. It also uses right hand or left hand rule to traverse the hole. It also uses a temporary void node in every bypass mode. If the rules of anchor nodes are violated then the void node will be deleted. EDGR is also extended to three-dimensional (3D) sensor networks to achieve an efficient energy-aware routing approach to detour routing holes.
4. Comparison of different routing hole bypassing techniques
Table 1 represents the comparison between the proposed hole handing techniques which are discussed in the previous section based on the mechanisms used to achieve fault forbearance with respect to routing hole problem, their characteristics and maintenance of states are highlighted for each of the approaches.
Table 1: Comparison of some routing hole bypass techniques
Technique | Maintenance of state | Mechanism used to fault forbearance with routing holes | Delivery of packets | States | Complexity | Overhead | Scalability | Optimality of path | Distribution of information |
GPSR |
Planer graph and geographic information | Greedy forwarding, face routing, Right hand rule and planer graph using GG and RNG graph |
Guaranty delivery |
Stateless |
Medium |
Medium |
Scalable |
Not optimal |
Localized |
Compass Routing-II, Face-I, Face-II |
Planar graph and Geographic information | Greedy forwarding, Face routing, Right hand rule and planer sub graph, Fall back mechanism |
Guaranty delivery |
Stateless |
Medium |
High |
Not Scalable |
Not optimal |
Localized |
GOAFR+
|
Planar graph and Geographic information | Greedy forwarding, Face routing, Right hand rule and planer graph |
Guaranty delivery |
Stateless |
Medium |
High |
Not Scalable |
Not optimal |
Localized |
Intermediate Node Forwarding |
Geographic information | Negative acknowledgment packets
|
No guarantee delivery |
Stateless |
Medium |
High |
Not Scalable |
Optimal |
Global |
GEAR |
Learned cost and estimated cost | Learned cost helps to identify alternative path and limited flooding in region of interest |
No guarantee delivery |
State free |
High |
Low |
Not scalable |
Not optimal |
Localized |
BOUNDHOLE |
Geographic information and holes boundary |
TENT rule to detect holes and maintains information of boundary |
Sometimes |
Stateless |
High |
Medium |
Scalable |
Not optimal |
Localized |
Curved Stick(CS) |
Geographic information |
Hop count reduction method |
No guarantee delivery |
Stateless |
High |
High |
Not scalable |
Not optimal |
Localized |
LRS |
Data aggregation on nodes |
Hop-election |
Sometimes |
State free |
Medium |
High |
Not scalable |
Optimal |
Localized |
EDGR |
Geographic information, Residual energy, energy consumption characteristics |
Anchor list, Right hand rule, left hand rule |
Guarantee delivery |
Stateless |
High |
High |
Scalable |
Optimal |
Localized |
6. Conclusion
As routing holes present a serious problem on the performance of the sensor network, it has to be resolved. Geographic routing approach is one of the commonly used schemes to overcome from the holes problem. In this survey paper we discussed about the basic principles of geographic routing for WSNs. We have presented a detailed review on different routing hole handing techniques available in literature and their possible strengths and short comes. Later these approaches were compared in a qualitative manner, using nine criteria from various perspectives.
References
[1] D. Torrieri et al., “Performance Comparisons of Geographic Routing Protocols in Mobile Ad Hoc Networks,” IEEE Trans.Commun., vol. 63, no. 11, 2015, pp. 4276–86.
[2] A. Mostefaoui et al., “Localized Routing Approach to Bypass Holes in Wireless Sensor Networks,” IEEE Trans. Comput., no. 63, no. 12, Dec. 2014, pp. 3053–65.
[3] C. Petrioli et al., “ALBA-R: Load-Balancing Geographic Routing Around Connectivity Holes in Wireless Sensor Networks,” IEEE Trans. Parallel Distrib. Sys., vol. 25, no. 3, Mar. 2014, pp. 529–39.
[4] C. Fraser et al., “A Survey of Geographical Routing in Wireless Ad-Hoc Networks,”IEEE Commun. Surveys & Tutorials,vol. 15, no. 2, 2013, pp. 621–53.
[5] Moyses M. Lima et al., " A Novel RSSI-based Algorithm for Detect and Bypass Routing Holes in Wireless Sensor Networks,"IEEE Symposium on Computers and Communications, 2018, pp. 00986-00989
[6] Vangili, et.al.", Detection of black hole attack in mobile ad-hoc networks using ant colony optimization–simulation analysis". Indian J. Sci. Technol. 8(13)-2015
[7] Das S., et.al., "Hole Detection in Wireless Sensor Network: A Review". In: Sa P., Bakshi S., Hatzilygeroudis I., Sahoo M. (eds) Recent Findings in Intelligent Computing Techniques. Advances in Intelligent Systems and Computing, vol 708. Springer, Singapore, 2018
[8] Fan Zhou et.al.,"Bypassing holes in sensor networks: Load -balance vs. latency" Ad Hoc Networks,Volume 61, June 2017, pp. 16-32
[9] Huy Vu, et al.,"An efficient geographic algorithm for routing in the proximity of a large hole in wireless sensor networks", Proceedings of the Seventh Symposium on Information and Communication Technology, 2016, pp. 286-293.
[10] Habib M. Ammari, "Investigating the Energy Sink-Hole Problem in Connected k-Covered Wireless Sensor Networks" IEEE transactions on computers, vol. 63, 2014 ,pp.1-14
[11] H.A.B.F.D. Oliveira, et al., "An enhanced location-free Greedy Forward algorithm with hole bypass capability in wireless sensor networks", Journal of Parallel and Distributed Computing, vol. 77, pp. 1-10, October 2015.
[12] M. Lima, et al.,"Geographic routing and hole bypass using long range sinks for wireless sensor networks", Ad Hoc Networks, vol. 67, Dec. 2017, pp. 1-10
[13] Jinwei Liu, et al., "Characterizing Data Deliverability of Greedy Routing in Wireless Sensor Networks", IEEE transactions on mobile computing, vol. 17, no. 3, march 2018 pp.543-559
[14] J. Ren et al., "Lifetime and energy hole evolution analysis in data-gathering wireless sensor networks", IEEE Trans. Ind. Informat., vol. 12, no. 2 Apr. 2016, , pp. 788-800,.
[15] K. Latif, et al., "On energy hole and coverage hole avoidance in underwater wireless sensor networks", IEEE Sensors J., vol. 16, no. 11, Nov. 2016, pp. 4431-4442
[16] D. Chen et al., “A Survey of Void Handling Techniques for Geographic Routing in Wireless Networks,” IEEE Commun. Surveys & Tutorials, vol. 9, no.1, 2007, pp. 50–67.
[17] Huang, H., et al., Coordinate-assisted routing approach to bypass routing holes in wireless sensor networks. IEEE Commun. Mag. 55(7), 2017, pp.180–185
[18] Q. Fang et al., “Locating and Bypassing Routing Holes in Sensor Networks,” IEEE INFOCOM ’04, Mar. 2004, pp. 7–11.
[19] B. Karp et al., “GPSR: Greedy Perimeter Stateless Routing for Wireless Networks,” ACM MobiCom ’00, Aug. 2000, pp. 243–54.
[20] Rehman, A et.al, "Sinkhole Attacks in Wireless Sensor Networks: A Survey" springer Wireless Pers Commun 2018, pp.1-23
[21] Ramazani, et.al.," Topological and combinatorial coverage hole detection in coordinate-free wireless sensor networks." Int. J. Sens. Netw. 21(1), 2016, pp. 40–52 .
[22] Wood, et.al "JAM: a jammed-area mapping service for sensor networks.", In: 24th IEEE Real Time System Symposium (RTSS’03),2003, pp. 286–298
[23] T. Amgoth, P. et al., "Coverage hole detection and restoration algorithm for wireless sensor networks", Peer-to-Peer Netw. Appl., vol. 10, no. 1, 2017, pp. 66-78
[24] Gurjinder Kaur et al., " Detection and Prevention of Blackhole Attacks in Wireless Sensor Networks". Springer, 2017, pp118–126
[25] Ali Benzerbadj, et al.,”Cross-Layer Greedy Position-Based Routing for Multihop Wireless Sensor Networks in a Real Environment,” Ad Hoc Networks (2018)
[26] H. Huang, et al., "Energy-Aware Dual-Path Geographic Routing to Bypass Routing Holes in Wireless Sensor Networks," in IEEE Transactions on Mobile Computing, vol. 17, no. 6, 1 June 2018, pp. 1339-1352
[27] Y.Yu, et al., "Geographiacal and energy aware routing: A recursive data dissemination protocol for wireless sensor networks." Technical Report UCLA/CSD-TR-01-0023,UCLA Computer Science Department ,2001
[28] Evangelos Kranakis et al., "Compass routing on geometric networks.", In Proceedings of the 11th Canadian Conference on Computational Geometry, 1999, pp 51.54.
[29]P. Bose et al.,”Routing with guaranteed delivery in ad hoc wirelessnetworks". Wireless Networks, 7:, 2001,pp. 609-616
[30] Fabian Kuhn, et al.,"Geometric ad-hoc routing: Of theory and practice". In 23rd ACM Symposium on Principles of Distributed Computing (PODC '03), July 2003.
[31] Douglas S. J., et al., "Location proxies and intermediate node forwarding for practical geographic forwarding". Technical Report MITLCS- TR-824, MIT Laboratory for Computer Science, June 2001.
[32] N. Sadeghzadeh-Nokhodberiz and J. Poshtan, "Distributed Interacting Multiple Filters for Fault Diagnosis of Navigation Sensors in a Robotic System," in IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 47, no. 7, pp. 1383-1393, July 2017.
[33] H. Phan, D. V. Dao, K. Nakamura, S. Dimitrijev and N. Nguyen, "The Piezoresistive Effect of SiC for MEMS Sensors at High Temperatures: A Review," in Journal of Microelectromechanical Systems, vol. 24, no. 6, pp. 1663-1677, Dec. 2015.
[34] M. Asgharpoor Salkuyeh and B. Abolhassani, "An Adaptive Multipath Geographic Routing for Video Transmission in Urban VANETs," in IEEE Transactions on Intelligent Transportation Systems, vol. 17, no. 10, pp. 2822-2831, Oct. 2016.
[35] L. Zhu, C. Li, B. Li, X. Wang and G. Mao, "Geographic Routing in Multilevel Scenarios of Vehicular Ad Hoc Networks," in IEEE Transactions on Vehicular Technology, vol. 65, no. 9, pp. 7740-7753, Sept. 2016.
[36] A. Awang, K. Husain, N. Kamel and S. Aïssa, "Routing in Vehicular Ad-hoc Networks: A Survey on Single- and Cross-Layer Design Techniques, and Perspectives," in IEEE Access, vol. 5, pp. 9497-9517, 2017.
[37] X. Jin, R. Zhang, J. Sun and Y. Zhang, "TIGHT: A Geographic Routing Protocol for Cognitive Radio Mobile Ad Hoc Networks," in IEEE Transactions on Wireless Communications, vol. 13, no. 8, pp. 4670-4681, Aug. 2014.
[38] W. Jung, J. Yim and Y. Ko, "QGeo: Q-Learning-Based Geographic Ad Hoc Routing Protocol for Unmanned Robotic Networks," in IEEE Communications Letters, vol. 21, no. 10, pp. 2258-2261, Oct. 2017.
[39] H. Zhou, S. Xia, M. Jin and H. Wu, "Localized and Precise Boundary Detection in 3-D Wireless Sensor Networks," in IEEE/ACM Transactions on Networking, vol. 23, no. 6, pp. 1742-1754, Dec. 2015.
[40] D. Zhang and E. Dong, "A Virtual Coordinate-Based Bypassing Void Routing for Wireless Sensor Networks," in IEEE Sensors Journal, vol. 15, no. 7, pp. 3853-3862, July 2015.
[41] R. W. L. Coutinho, A. Boukerche, L. F. M. Vieira and A. A. F. Loureiro, "Geographic and Opportunistic Routing for Underwater Sensor Networks," in IEEE Transactions on Computers, vol. 65, no. 2, pp. 548-561, 1 Feb. 2016.
[42] M. Z. Hasan, F. Al-Turjman and H. Al-Rizzo, "Analysis of Cross-Layer Design of Quality-of-Service Forward Geographic Wireless Sensor Network Routing Strategies in Green Internet of Things," in IEEE Access, vol. 6, pp. 20371-20389, 2018.
[43] C. Zhu, L. T. Yang, L. Shu, V. C. M. Leung, J. J. P. C. Rodrigues and L. Wang, "Sleep Scheduling for Geographic Routing in Duty-Cycled Mobile Sensor Networks," in IEEE Transactions on Industrial Electronics, vol. 61, no. 11, pp. 6346-6355, Nov. 2014.
[44] D. Zhang and E. Dong, "An Efficient Bypassing Void Routing Protocol Based on Virtual Coordinate for WSNs," in IEEE Communications Letters, vol. 19, no. 4, pp. 653-656, April 2015.
[45] L. Shu, M. Mukherjee, L. Hu, N. Bergmann and C. Zhu, "Geographic Routing in Duty-Cycled Industrial Wireless Sensor Networks With Radio Irregularity," in IEEE Access, vol. 4, pp. 9043-9052, 2016.
[46] H. Huang, H. Yin, Y. Luo, X. Zhang, G. Min and Q. Fan, "Three-dimensional geographic routing in wireless mobile ad hoc and sensor networks," in IEEE Network, vol. 30, no. 2, pp. 82-90, March-April 2016.
[47] H. P. Gupta, S. V. Rao, A. K. Yadav and T. Dutta, "Geographic Routing in Clustered Wireless Sensor Networks Among Obstacles," in IEEE Sensors Journal, vol. 15, no. 5, pp. 2984-2992, May 2015.
[48] A. M. Popescu, N. Salman and A. H. Kemp, "Energy Efficient Geographic Routing Robust Against Location Errors," in IEEE Sensors Journal, vol. 14, no. 6, pp. 1944-1951, June 2014.
[49] L. Cheng, J. Niu, J. Cao, S. K. Das and Y. Gu, "QoS Aware Geographic Opportunistic Routing in Wireless Sensor Networks," in IEEE Transactions on Parallel and Distributed Systems, vol. 25, no. 7, pp. 1864-1875, July 2014.
[50] S. M. Ghoreyshi, A. Shahrabi and T. Boutaleb, "Void-Handling Techniques for Routing Protocols in Underwater Sensor Networks: Survey and Challenges," in IEEE Communications Surveys & Tutorials, vol. 19, no. 2, pp. 800-827, Secondquarter 2017.
[51] B. Khalifa, Z. Al Aghbari, A. M. Khedr and J. H. Abawajy, "Coverage Hole Repair in WSNs Using Cascaded Neighbor Intervention," in IEEE Sensors Journal, vol. 17, no. 21, pp. 7209-7216, 1 Nov.1, 2017.
[52] F. Yan, A. Vergne, P. Martins and L. Decreusefond, "Homology-Based Distributed Coverage Hole Detection in Wireless Sensor Networks," in IEEE/ACM Transactions on Networking, vol. 23, no. 6, pp. 1705-1718, Dec. 2015.
