Routing protocols for ad hoc networks

Ad hoc networks
OUTLINE:
• Introduction

• Classification

• Proactive routing protocols

• Reactive routing protocols

• Hybrid routing protocols

• Hierarchial routing protocols

• Power-aware routing protocols

2
1. Introduction
W h a t we have in ad hoc environment:
• dynamically changing topology;
• absence of fixed infrastructure and centralized administration;
• bandwidth constrained wireless links;
• energy-constrained nodes.

Link loss is one of the biggest problem for routing.

3
1. Challenges of routing protocols in ad hoc networks
T h e following are the main challenges:
• Movement of nodes:
– Path breaks;
– Partitioning of a network;
– Inability to use protocols developed for fixed network.

• Bandwidth is a scarce resource;

– Inability to have full information about topology;
– Control overhead must be minimized.

• Shared broadcast radio channel:

– Nodes compete for sending packets;
– Collisions.

• Erroneous transmission medium:

– Loss of routing packets.

4
2. Design goals
Goals that must be met:
• must be scalable;

• must be fully distributed, no central coordination;

• must be adaptive to topology changes caused by movement of nodes;

• route computation and maintenance must involve a minimum number of nodes;

• must be localized, global exchange involves a huge overhead;

• must be loop-free;

• must effectively avoid stale routes;

• must converge to optimal routes very fast;

• must optimally use the scare resources: bandwidth, battery power, memory, computing;

• should provide QoS guarantees to support time-sensitive traffic.

5
2. Classification of routing protocols
Ro u ti n g protocols for ad-hoc wireless networks can be classifi ed based on:
• routing information update mechanism;

• usage of temporal information (e.g. cached routes);

• usage of topology information;

• usage of specific resources (e.g. G PS) .

6
Based on routi ng informati on update mechanism
• Proactive (table-driven) routing protocols;
• Reactive (on-demand) routing protocols;
• Hybrid protocols.

Based on usage of temporal informati on

• Based on past temporal information;
• Based on future temporal information.

Based on the routi ng topology

• Flat topology routing protocols:
• Hierarchial topology routing protocols:

Ro u ti n g based on uti lizati on of specifi c resources:

• Power-aware routing;
• Geographical information assisted routing.

7
 Routing protocols for ad-hoc networks

Routing information update Temporal information for routing

Proactive Reactive Hybrid Past history Predictions

- DSDV; - DSR; - CEDAR; - DSDV; - FORP;

- WRP; - AODV; - ZRP; - WRP; - RABR;
- CGSR; - ABR; - ZHLS. - STAR; - LBR.
- STAR; - SSA; - DSR;
- OLSR; - FORP; - AODV;
- FSR; - PLBR. - FSR;
- HSR; - HSR;
- GSR. - GSR.

Classification of routing protocols

8
Destination-Sequenced Distance-Vector Routing(DSDV)
Structured Transparent Accessible Reproducible(STAR)
cluster-head gateway switch routing(CGSR)
Optimized Link State Routing Protocol(OLSR)
Fisheye State Routing (FSR)
hierarchical routing protocol(HSR)
Global state routing(GSR)
Global state routing(GSR)
Ad-hoc On-demand Distance Vector(AODV)
Associativity-based routing (ABR)Signal Stability based Adaptive(SSA)
Flow-Oriented Routing Protocol
Preferred Link-based Routing(PLBR)
Core Extraction Distributed Ad Hoc Routing Protocol
Zone Routing Protocol
Zone-based Hierarchical Link State routing
Flow oriented routing protocol
Route-lifetime assessment based routing
Load balancing routing
Location-aided routing
 Routing protocols for ad-hoc networks

Utilization of specific resource Topology information

Power-aware Geographical Flooding Flat Hierarchial

- PAR. - LAR. - DSR; - CGSR;

- AODV; - FSR;
- ABR; - HSR.
- SSA;
Proactive Reactive - FORP;
- PLBR.
- OLSR. - PLBR.

Classification of routing protocols, contd…

1
0
Ad hoc networks D.Moltchanov, T U T, 2009

3. Proactive routing protocols

These are table-driven protocols.
We consider:
• destination sequenced distance vector routing protocol (DSDV);

• wireless routing protocol ( W R P ) ;

• cluster head gateway routing protocol ( C G C R ) .

• source-tree adaptive routing protocol ( STA R ) ;

C o m m o n advantages and shortcoming of these protocols:

• + : low delay of route setup process: all routes are immediately available;
• − : high bandwidth requirements: updates due to link loss leads to high control overhead;

• − : low scalability: control overhead is proportional to the number of nodes;

• − : high storage requirements: whole table must be in memory.

1. Destination sequenced distance vector routing protocol ( D S DV )
Modifi cati on of the Bellman-Ford algorithm where each node maintains:
• the shortest path to destination;

• the first node on this shortest path.

T h i s protocol is characterized by the following:

• routes to destination are readily available at each node in the routing table (RT);
• R Ts are exchanged between neighbors at regular intervals;

• R Ts are also exchanged when significant changes in local topology are observed by a node.

R T updates can be of two types:

• incremental updates:
– take place when a node does not observe significant changes in a local topology;

• full dumps:
– take place when significant changes of local topology are observed;
 3 Dest Next Dist
1
6 3 2
6
5 3 2
4 4 2 2
2 5

Example of routing table in D S DV.

T h e reconfi gurati on of path (used for ongoing data transfer) is done as follows:
• the end node of the broken link sends a table update message with:
– broken link’s weight assigned to infinity;
– sequence number greater than the stored sequence number for that destination.

• each node re-sends this message to its neighbors to propagate the broken link to the
network;
• even sequence number is generated by end node, odd – by all other nodes.

Note: single link break leads to the propagation of R T updates through the whole network!
 3 3 Dest Next Dist
1 1
6 3 3
6
5 3 2
4 4 4 2 2
2 5 2 5 6

Route update in D S DV.

Ro u te maintenance in D S D V is performed as follows:

• when a neighbor node perceives a link break (node 3):
– it sets all routes through broken link to ∞;
– broadcasts its routing table.

• node 5 receives update message, it informs neighbors about the shortest distance to node 6;

• this information is propagated through the network and all node updates their RTs;

• node 1 may now sends their packets through route 1 − 3 − 5 − 6 instead of 1 − 3 − 6.

2. Wireless routing protocol ( W R P )
E a c h node maintains the following:
• Distance table ( D T ) containing network
view of the neighbors of the node:
– distance and predecessor node for all
destinations as seen by each neighbor.
• Routi ng table ( R T ) containing view of the
network for all known destinations
including:
– the shortest distance to destinations;
– the predecessor node;
– the successor node;
– flag indicating the status of the path
(correct, loop, null).
• Link cost table ( L C T ) containing cost-
related information including:
– number of hops to reach destination
(cost of the broken link is ∞);
– number of update periods passed from
the last successful update of the link.
Routing entries at each node for destination 7: Dest Next Pred Cost
2 4 7 7 7 0
1
6 3 5 8
3 6
3 5 7 5 3
4 1
Source 7
2 4 2 4 7 4 2
2 5
3 5 5 4
3 2 3
2 7 2 3
7 1 3 5 6
Destination

Routi ng table in W R P.

B reak: detected by number of update periods missed since successful

transmission:
• each update message contains a list of updates;
• a node marks each node in R T that has to acknowledge update message it transmitted;
• once the counter of M R L reaches zero:
– entries in update message for which no acknowledgement received are to be
retransmitted;
– update message is deleted.
Ad hoc networks D.Moltchanov, T U T, 2009
W h e n a node detects a link break:
• it sends an update message to its neighbors with link cost set to ∞;
• all affected node update their routes when they receive update message;
• the node that initiated the update message finds an alternative route from its D T;
• it updates its R T and sends update message to its neighbors;
• nodes update its R Ts if this received route is better than they have, otherwise they discard it.
Dest Next Pred Cost Dest Next Pred Cost
7 7 7 0 2 3 4 7 7 7 0
1
6 3 5 8 Source 6 3 4 9
3 1 6
5 7 5 3 4 5 7 5 4 4 4
4 7 4 2 2 4 4 7 4 2
2
3 5 5 4 2 3 4 4 5
3 2
2 7 2 3 3 2 7 2 3
1 3 5 6 7 1 2 2 7
Destination

Route maintenance in W R P.
Ad hoc networks D.Moltchanov, T U T, 2009

3. Cluster head gateway switch routing protocol ( C G S R )

I t is characterized by the following:
• nodes are organized into clusters, each having an elected cluster-head;
• cluster head provides a coordination within its transmission range (single hop);
• token-based scheduling is used within a cluster for sharing bandwidth between nodes;
• all communications pass through the cluster head;
• communication between cluster is done using the common nodes (gateways with two interfaces).

Abstract representation of routing in G G S R .

Ad hoc networks D.Moltchanov, T U T, 2009
Tw o tables are used in C G S R :
• Cluster member table containing the destination cluster head for every node;

• Routi ng table containing the next-hop node for every destination cluster.

T h e protocol operates as follows:

• a node obtains a token from its cluster head;

• if this node has a packet to transmit, it determines the destination cluster head and next-hop;

• routed packet goes as follows:

– a − H 1 − G 1 − H 2 − G 2 − · · · − H i − G i − · · · − G n − H n − b: where
– G i is the gateway i;
– H i is the cluster head i.

Ro u te reconfi gurati on happens when:

• there is a change in cluster heads;

• stale entries are in cluster member table or routing table.

 Gateway node
10

Cluster head
2

11
1
Source and destination

Destination
6

Source
5 9 12

14
7
15
8

Route path in C G S R .
4. Source-tree adaptive routing protocol ( S TA R )
T h e re are two protocols with diff erent aims:
• Least overhead routing approach ( LO R A ) :
– minimize control overhead irrespective of optimality;
• Optimum routing approach ( O R A ) :
– provide optimal routes irrespective of the control overhead;

T h e S T A R protocol operates as follows:

• each node is required to:
– send an update message to its neighbors during initialization;
– send update messages about new destinations, chances of routing loops, costs of paths.
• every node broadcasts its source-tree information:
– wireless links used by the node in its preferred path to destinations.
• every node builds its partial graph of topology based on:
– its adjacent links with neighbors, source-tree broadcasts by neighbors.
 7 6 7 6

8 3 8 3
5 5
1 4 1 4

2 2

Local topology construction.

T h e following method is used to fi nd paths:

• if a node 1 wishes to send data to node N and does not have path in its source tree:
– it sends update message to all neighbors and indicates that there is no path to
N;
– neighbor that have the path, responses with update messages;
– node 1 updates its source tree and may begin transmission.
4. Reactive routing protocols
These protocols find paths to destination only when needed (on-demand) to
transmit a packet.
We consider:
• Dynamic source routing protocol (DSR);
• Ad hoc on-demand distance vector routing protocol (AODV);
• Location aided routing (LAR);
• Associativity-based routing (ABR);
• Signal stability-based adaptive routing protocol (SSA);
These protocols have the following advantages and shortcomings:
• −: high delay of route setup process: routes are established on-demand;
• + : small control overhead: no route updates;
• −: low scalability: no route updates;
• −: low storage requirements: only needed routes are in cache.
1. Dynamic source routing protocol
This is a source-based routing protocol.
T h e difference between D S R and other on-demand routi ng
protocols is:
• on-demand protocols periodically exchange the so-called beacon (hello)
packets:
– hello packets are used to inform neighbors about existence of the node.
• D S R does not use hello packets.
T h e basic approach of this protocol is as follows:
• during route contraction D S R floods a RouteRequest packets in the
network;
• intermediate nodes forward RouteRequest if it is not redundant;
• destination node replies with RouteReply;
• the RouteReply packet contains the path traversed by RouteRequest
packet;
• the receiver responds only if this is a first RouteRequest (not duplicate).
T h e D S R protocol uses the sequence numbers:
• RouteRequest packet carries the path traversed and the sequence number;
• the sequence numbers are used to prevent loop formation and nodes check it.
T h e D S R also uses route cache in each node:
• if node has a route in the cache, this route is used.

8
R
12
7

9
10

13 6

4
5
2

14
S 16

Route establishment in D S R .
2. Ad hoc on-demand distance vector routing protocol
T h e major differences between A O D V and D S R are as follows:
• in D S R a data packet carries the complete path to be traversed;
• in AODV nodes store the next hop information (hop-by-hop routing) for
each data flow.
T h e RouteRequest packet in A O D V carries the following
informati on:
• the source identifier (SrcID): this identifies the source;
• the destination identifier (DestID): this identifies the destination to which
the route is required;
• the source sequence number (SrcSeqNum);
• the destination sequence number (DestSeqNum): indicates the freshness
of the route.
• the broadcast identifier (BcastID): is used to discard multiple copies of the
same RouteRequest.
• the time to live ( T T L ) : this is used to not allow loops.
T h e A O D V protocol performs as follows:
• when a node does not have a valid route to destination a RouteRequest is
forwarded;
• when intermediate node receives a RouteRequest packet two cases are
possible:
– if it does not have a valid route to destination, the node forwards it;
– if it has a valid route, the node prepares a RouteReply message:
• if the RouteRequest is received multiple times, the duplicate copies are
discarded:
– are determined comparing BcastID-SrcID pairs.
• when RouteRequest is forwarded, the address of previous node and its
BcastID are stored;
– are needed to forward packets to the source.
• if RouteReply is not received before a time expires, this entry is deleted;
• either destination node or intermediate node responses with valid route;
• when RouteRequest is forwarded back, the address of previous node and
its BcastID are stored;
 R RouteRequest:
17 - DestSeqNum:5,
14 - SrcSeqNum: 2
13 Node 8:
11 - DestSeqNum 3;
12
Node 9:
16 - DestSeqNum 7;
8
6
9 7

10 1
2
5

S 4

Route establishment in A O DV.

 I n case of the link break:
• end-node are notified by unsolicited RouteReply with hop count set to ∞;
• end-node deletes entries and establishes a new path using new BcastID;
• link status is observed using the link-level beacons or link-level A C K s .
D
17
14
13

11 12
16
8
6
9 7

10 1
2
5

S 4

Route maintenance in A O DV.

3. Location aided routing
Basic of L A R :
• uses the location information (assumes the availability of GPS);
• reactive (on-demand) protocol.
L A R designates two zones for selective forwarding of control packets:
• ExpectedZone:
This is a geographical zone in which the location of the terminal is predicted based on:
– location of the terminal in the past;
– mobility information of the terminal.
There is no info about previous location of the terminal the whole network is the
ExpectedZone.
• RequestZone:
This is a geographical zone within which control packets are allowed to propagate:
– this area is determined by the sender of the data packet;
– control packets are forwarded by node within a RequestZone only;
– if the node is not found using the first RequestZone, the size of RequestZone is
increased.
 1
2 5

8 Dest.
3 10
9
6

3 Source 7
4

Example of ExpectedZone and RequestZone.

Nodes decide whether to forward or discard packets based on two algorithms:

• L A R type 1;

• L A R type 2.
L A R type 1 algorithm works as follows:
• the sender explicitly specifies the RequestZone in the RouteRequest packet;
• the RequestZone is the smallest rectangle that includes the source and the
ExpectedZone;
• when the node is in ExpectedZone, the RequestZone is reduced to the
ExpectedZone;
• if the ReouteRequest packet is received by the node within RequestZone, it
forwards it.
1
L A R type 2 algorithm operates as follows:
• the sender includes the distance to the source in the RouteRequest packet;
• intermediate nodes compute the distance to the destination:
– if this distance is less than the distance between source and destination
packet is forwarded;
– otherwise the packet is discarded.
• distance in the packet is updated at every node with lower distance to
destination.
1 14

2 5 16
12
8 Dest. 17
3 10 13
9
6
18
11 Source 7 15
4

Routi ng procedure in L A R type 2.

4. Associativity-based routing
It is characterized by the following:
• on-demand beacon-based protocol;
• routes are selected based on temporal stability of wireless links:
• to determine temporal stability, each node maintains the count of its
neighbors’ beacons.
T h e protocol operates as follows:
• source node floods the RouteRequest packet, all intermediate nodes
forward this packet;
• RouteRequest packet carries the following:
– the path it has traversed;
– the beacon count for corresponding node in the path.
• when the first RouteRequest reaches the destination, the destination:
– waits for RouteSelect time to receive multiple copies of
RouteRequest;
– selects the path that has the maximum number of stable links;
– replies to the source with RouteReply packet.
 1 14
Dest.
2 5 16
12
8
3 10 13
9
6

11 4 7 15
Source
4 - 7 - 10 - 5 - 14;
4 - 7 - 6 - 12 - 14;
4 - 9 - 8 - 1 - 5 - 14...

Routi ng in A B R .
 If the link break occurs:
• the node closer to the source initiates a local link repair as follows:
– broadcasts locally route repair packet (local query ( LQ ) ) with limited T T L (e.g., 3);
• if this node fails to repair, then the next node closer to destination initiates a route repair;
• if nodes constituting a half of pass of the route fail to repair, the source is informed.

1 14
Dest.
2 5 16
12
8
3 10 13
9
6

11 4 7 15
Source

Local route repair in A B R .

5. Signal stability-based adaptive routing protocol
T h i s protocol is characterized by the following:
• on-demand beacon-based protocol;

• routes are selected based on temporal stability of wireless links:

• based on temporal stability, each links is classified to:

– stable
– unstable

• to determine temporal stability, each node measures the signal strength of beacons.

T h e whole protocols consist of the following two parts:

• dynamic routing protocol ( D R P ) :
– D R P maintains the routing table interacting with D R P s on other hosts.

• forwarding protocol (FP) :

– is responsible for forwarding of packets to destination.
I n each node the signal stability table is maintained containing:
• beacon count and signal strength of these beacons;
– if the signal strength is strong for past few beacons the link is stable;
– if the signal strength is weak for past few beacons the link is unstable.
T h e protocol operates as follows:
• if no route in cache, the node floods the RouteRequest packet;
• RouteRequest packet carries the path it has traversed;
• if the intermediate node receives the RouteRequest via stable link it forwards it;
• if the intermediate node receives the RouteRequest via unstable link it drops it;
• when the first RouteRequest reaches the destination, the destination:
– waits for RouteSelect time to receive multiple copies of RouteRequest;
– selects the path that is most stable;
∗ if two or more paths are equal in stability, the shortest path is selected;
∗ if two or more shortest paths are available, random path among them is selected.
– replies to the source with RouteReply packet.
 1 14
Dest.
2 5 16
12
8
3 10 13
9
6

11 4 7 15
Source

4 - 7 - 10 - 5 - 14;
4 - 7 - 10 - 5 - 12 - 14.

Routi ng in SSA.
 I n case of the link break:
• end-nodes are informed and they try to establish new stable route;
• of no stable routes are available, the restriction of stable links is removed.

1 14
Dest.
2 5 16
12
8
3 10 13
9
6

11 4 7 15
Source

4 - 7 - 10 - 5 - 14;
4 - 7 - 10 - 5 - 12 - 14.

Route repair in SSA.

6. Flow-oriented routing protocol
A i m is on supporti ng real-ti me traffi c using a predicti ve multi -hop-handoff feature:
• Classic protocols (e.g., D S R , A O D V etc.):
– route repair is initiated when intermediate node detects the link breaks;
– it causes delay, losses: low QoS is provided in result.

• F O R P uses the predictive mechanism to estimate link expiration time ( L E T ) :

– it is based on location, mobility and transmission range of nodes involved in forwarding;
– the minimum of L E T determines the route expiration time ( R E T ) ;
– it is assumed that G P S is used to make prediction of L E T.

N e w shortcomings:
• − : devises are expensive;

• − : can only operate at the open air (due to G PS) .

T h e protocol operates as follows:
• if no route in cache, the destination floods the Flow-REQ packet carrying:
– information regarding the source and destination nodes;
– flow identification number (sequence number) that is unique for every session.

• when neighbor receives the Flow-REQ packet:

– to avoid looping checks if the sequence number if higher than that previously used;
– if so, this node appends L E T and its address in the packet, and forwards it;
– if not, the packet is discarded.

• when the destination receives the packet:

– the packet has a path it has traversed and L E T associated with each wireless link;
– if R E T is acceptable, it originates the Flow-SETUP packet.

• when the source receives Flow-SETUP packet, it begins the transmission of packets.
 1 Dest. 14

2 14 9 5 16
12
8 8 5
3 10 13
9
10 6
7
11 17 7 15
4
Source

4 - 7 - 10 - 5: RET: 8
4 - 9 - 8 - 1 - 5: RET: 5

Route establishment in F O R P.
 T h e L E T of the link is esti mated as follows:
−(pq + rs) + (p 2 + r 2 )T X2 − (ps − qr)2
LETA B = ,
p2 + q2
p = VA cos T A − VB cos T B , q = XA − XB
r = VA sin T A − VB sin T B , s = YA − YB , (1)
• A and B are nodes with transmission range T X ;
• V A and V B are velocities of nodes;
• T A and T B are angles as shown below:

TA TB

A: (XA,YA) B: (XB,YB)

Motion angles in FO R M .
F O R P uses proacti ve route maintenance using available R E T :
• when the destination determines that the break is about to occur it sends Flow-HANDOFF;
• Fl ow-HAN DO FF propagates in the network similarly to Flow-REQ ;
• when many Fl ow-HAN DO FF are received at the source new path with highest R E T is chosen.

Dest. 14 5 - 10 - 7 - 4: RET: 5
1
5 - 1 - 8 - 9 - 4: RET: 8
2 5 16 5 - 1 - 3 - 4: RET: 3
6 14 9
12 5 - 6 - 7 - 4: RET: 7
8 8 5
5 7
3 3 10 13
9
10 6
7 9
11 17 7 15
4
Source

Route repair in F O R P.
5. Hybrid routing protocols
These protocols maintain topology information up to m hops in tables.
We consider:
• Zone routing protocol ( Z R P ) ;

• Zone-based hierarchial link state routing protocol ( ZHL S) ;

W h a t are inherent shortcomings and advantages:

• + : fast link establishment;

• + : less overhead as compared to table-driven and reactive protocols.

• − : high storage and processing requirements as compared to reactive protocols.

Note: a compromise between proactive and reactive protocols.

1. Zone routing protocol ( Z R P )
This protocols uses a combinati on of proacti ve and reacti ve
routi ng protocols:
• proacti ve: in the neighborhood of r hops: Intra-zone routing protocol
(IARP);
• reacti ve: outside this zone: Inter-zone routing protocol (IERP).

14 r=2
1 Dest.
r=1
2 5 16
12
8
3 10 13
9
6

11 4 7 15
Source

For r = 2: 5, 6, 7, 9 are interior nodes

14, 12, 15, 4, 2, 8, 1 are paripheral nodes Routing within a zone toplogy information is
exchanged using route update packets.

Zones in Z R P for node 10.

T h e protocol operates as follows:
• if the destination is within the zone, the source sends packets directly;
• if not, the destination sends RouteRequest to peripheral nodes;
• if any peripheral node, has the destination in its zone it replies with
RouteReply;
• if not, peripheral nodes sends RouteRequest to their peripheral nodes
and so on;
• if multiple RouteReply are received the best is chosen based on some
metric.
If the broken link is detected:
• intermediate node repairs the link locally bypassing it (proactive
routing!!!);
• end nodes are informed;
• sub-optimal pass but very quick procedure;
• after several local reconfiguration, the source initiates global pass
finding to find optimal.
 1 14

2 5 16
12
8
Source
3 10 13
9
6

11 4 7 15
Dest.

Routi ng in Z R P with r = 1.
2. Zone-based hierarchial link state routing protocol
Z H L S is characterized by the following:
• use of geographical location of nodes to determine the non-overlapping zones;
• hierarchial addressing with zone ID and node ID is used;
• each node requires the location information based on which its zone is obtained;
• topology information is maintained in every node inside this zone;
• for regions outside the zone, zone connectivity information is maintained;
T h e Z H L S uses:
• proacti ve routi ng is used inside zone;
• reacti ve routi ng is used outside zone;
Note: Z H L S requires GP S or similar service to identify itself with a certain sone.
Zones: coverage of the single node, application scenario, mobility of nodes, network
size.
T h e protocol operates as follows:
• each node builds a one-hop node-level topology;
• this one-hop topology is propagated to other nodes in its zone using packet
containing:
– IDs of all zones in the zone, node ID, and zone IDs of all other nodes.
• nodes that receive responses from nodes belonging to other zones are gateway
nodes;
• all traffic between zones is transmitted via gateway nodes;
• once node-level topology is built, nodes obtain zone-level topology sending
packets via gates;
• if the destination is in the zone, packets are forwarded directly;
• if no, the source sends location request packet to every zone via gateways;
• every gateway node checks for destination in its routing table and replies with
ReouteReply.
 1 14
D C 2 B A 16
5 12
8
3 10 13
9
6

11 H G 7 F E
4 15

Z H L S zones.

T h e repair of broken links is as follows:

• source is notified about the link failures;
• if there are multiple gateways with the required zone, packet if forwarded via one of those;
• if no multiple gateways, packets are forwarded to other zones and then to the required zone.
6. Hierarchial routing protocols
These protocols introduce hierarchy in the network to achieve the following benefits:
• reduction in the size of routing tables;

• better scalability.

1. Hierarchial state routing protocol

H S R is characterized by the following:
• H S R uses multi-clustering to enhance resource allocation and management;

• H S R defines different levels of clusters;

• at every level leader is elected;

• the first level is made of single-hop clusters (physical clustering);

• the next level is comprised of leaders of clusters.

 Level 2

Level 1

Level 0 16
5 12
2
20 8
18
3 10 13
9
6
17
11 7
4 15 19

Topology example in H S R .
A t the physical layer nodes are classifi ed into:
• cluster heads; belong to a single cluster and elected as a cluster head;

• gateway nodes; belong to two or more clusters;

• normal nodes: belong to a single cluster.

C l uster heads at level 0 (physical level) could be responsible for:

• slot/frequency/code allocation to utilize spectrum more efficiently;

• call admission control from normal member nodes;

• scheduling of packets for transmission;

• exchange of routing information;

• handling route breaks.

Gateway nodes are responsible for:

• forwarding of packets between different clusters (cluster heads).
 T h e following routi ng responsibiliti es are assigned to nodes in H S R :
• every node maintains information about the status of links with its neighbors:
– this information is broadcasted within the cluster at regular intervals.
• cluster heads exchange the topology and link state information at any level:
– this is done via multiple hops using the gateway nodes.
• the path between two cluster head involves multiple links is called the virtual link:
– this is: head - gateway - head - gateway etc.
• every node knows the exact hierarchial topology information:
– after obtaining the information the cluster head floods it to lower level.
• to route packets hierarchial addressing is used consisting of;
– hierarchial I D ( H I D ) : sequence of cluster headers’ IDs from higher level;
–node I D : similar to unique M A C address.
Address of node 12: (9,12,12).
• From 12 to 3: 12 - 9 - 3 over multiple links.
2. Fisheye state routing protocol
Generalizati on of the G S R protocol where the following property is introduced:
• accurate information about nodes in local topology;

• not so accurate information about node that are far away.

W h y is it needed:
• complexity proactive routing: size of the network, mobility of node;

• reactive routing: + number of connections.

W h a t is the basis:
• a node exchanges the routing information only with neighbors at periodic intervals:
– trade-off between link-state (topology exchanges) and distance vector (link-level info).

• the complete topology information is maintained at each nodes;

• different update frequencies for different scopes one-hop/two-hop/... scopes:

– one-hop – highest freq., two-hop less freq. etc.: decrease of the message size.
 16
5 12
2
20 8
18
3 10 13
9
6
17
11 7
4 15 19

Figure 29: One-hop and two-hop scopes of the node.

2 Dest. Neigh. Hops

1 2,3 0
1
2 1 1 one-hop neighbors
3 1,4 1
4 3 4 3 2 two-hop neighbors

Example of topology information in F S R .

Ad hoc networks D.Moltchanov, T U T, 2009

7. Power-aware routing protocols

T h e following metrics can be taken into account on route selecti on procedure:
• M i n i m al energy consumpti on per a packet:
This metric involves a number of nodes from source to destination.
– + : uniform consumption of power throughout the network;
• M axi m i ze the network connecti vity:
To balance the load between the cut-sets (those nodes removal of which causes partitions).
– − : difficult to achieve due to variable traffic origination.
• M i n i m u m variance in node power levels:
To distribute load such that power consumption pattern remains uniform across nodes.
– + : nearly optimal performance is achieved by routing a packet to least loaded next-hop.
• M i n i m u m cost per a packet:
Cost as a function of the battery charge (less energy – more cost) and use it as a metric.
– + : easy to compute (battery discharge patterns are available);
– + : this metric handles congestions in the network.
Next Generation Network (NGN)
 CONTENTS
• INTRODUCTION
• EXISTING NETWORKS & THEIR DISADVANTAGES
• NGN : DEFINATION
• CHARACTERSTICS OF NGN
• NGN ARCHITECTURE
• NGN IN INDIA
• CONCLUSION
 INTRODUCTION
• Today ,telephony,the internet,and the cellular
mobile networks continue to be different
domains,each has its own protocols and
services.
• Telecommunication industry is changing at a
rapid pace. This change in the industry is
basically driven by demand of new services from
subscriber's side and urge to reduce CAPEX
(Capital Expenditure) and OPEX (Operational
Expenditure) from carrier side.
• NGN will be the foundation for creation of a new
range of multimedia applications that takes full
advantage of characterstics of broadband
network .
 EXISTING NETWORK
Presently telecommunications system employs
one of the three types of the following networks:
1) PSTN: Public Switch Telephone Network was
basically developed and engineered for giving
voice connectivity to the wire line subscribers.
2) PLMN (Public Land Mobile Network): PLMN
has been developed to provide voice services for
wireless subscribers
3) Data Network: This network was basically

designed for accessing remote files and servers

for defense people and universities but now a
days nobody can think of living without data
network services. The basic and most popular
application of data networks is Internet. Other
applications include E-commerce, online banking,
online gaming, E-shopping
 DISADVANTAGES
• Slow to develop new technologies
• Not compatible for IP platform
• No convergence features
• High CAPEX & OPEX due to maintenance of
different networks for different services.
• Large Power, Cooling requirements.
 NGN : DEFINATION

• A Next Generation Network

(NGN) is a packet-based network
able to provide
Telecommunication Services to
users and able to make use of
multiple broadband, QoS
enabled transport technologies
and in which service-related
functions are independent of the
underlying transport-related
technologies .
• NGN’s are commonly built around the internet protocol, & therefore the term “ALL IP” is also
used to describe them.
CHARACTERSTICS OF NGN
1) Packet based Transfer
• As per the new 4G and 5G technologies the
NGN is also a packet technology. The packet-
based networks provide for a higher speed and a
bandwidth efficient system.
• With the increase in number of users and the
need for high speed technology at the same
time, packet based systems are a boon to the
service providers
2) Decoupling of service provision from
transport-
• In an NGN,this means that whenever a provider
wants to enable a new service, they can do so
by defining it directly at the service layer without
considering the transport layer i.e. services are
independent of transport details.
3) Support for a Wide range of Services

• New and advanced devices which combine all

the existing services and devices into one which
serves for Customer satisfaction and also makes
it easy for the service providers at the same
time.
4) GENERALISED MOBILITY:
 ARCHITECTURE OF NGN
• A basic architecture was defined comprising
network elements needed for the provision of
traditional telephony services.
• The Next Generation Network architecture is
based on four layers: ACCESS LAYER , CORE
LAYER, CONTROL LAYER and SERVICE
LAYER .
ARCHITECTURE DIAGRAM
 NGN IN INDIA
• National Centre for Next Generation Network
(NCNGN) Circle has been created by the
Management Committee of BSNL Board in its
168th meeting held on 26th day of June 2013.
The NCNGN Circle with its headquarter at New
Delhi would be responsible for NGN project
coordination during installation and
commissioning.
NGN
IN INDIA
BY BSNL
 CONCLUSION
• The creation of the NGN is no overnight
transformation, but it is an evolution that is
already underway and gathering pace.
• With the fast paced modern lifestyle and growing
need for data access, the Next Generation
Networks are being designed which combine all
forms of data access into one. The NGN system
works like a black box which encapsulates all the
services into one.
• The NGN network is highly advantageous due
to this facility, for both customers and service
providers.
Multi Protocol Label Switching(MPLS)

• Labeled Packets and LSP

• Label Distribution
• RSVP-TE for MPLS
• Traffic Engineering Extensions to OSPF and IS-IS
Generalized MPLS

• GMPLS Labels 627

• Label Stacking and Hierarchical LSPs: MPLS/GMPLS 628
• RSVP-TE for GMPLS 629
• Routing Protocols in GMPLS 630
• Control and Data Path Separation and Link Management Protocol
Routing and Traffic Engineering with MPLS

• Traffic Engineering of IP/MPLS Networks

A Brisk Walk Back in History
MPLS-Based Approach for Traffic Engineering
• VPN Traffic Engineering
Problem Illustration: Layer 3 VPN
LSP Path Determination: Constrained Shortest Path Approach
LSP Path Determination: Network Flow Modeling Approach
• Layer 2 VPN Traffic Engineering
• Observations and General Modeling Framework
• Routing/Traffic Engineering for Voice Over MPLS
PSTN Call Routing Using the Internet

• Conceptual Requirement
• VoIP Adapter Functionality
• Addressing and Routing
• Service Observations
• Traffic Engineering
 PSTN and the Internet

Introduction of VoIP requires hybrid integration of the PSTN and the Internet.
There are three key environments that we will discuss in detail in this chapter.
They are broadly referred to as VoIP environments and can be briefly stated as
follows:
• A call originating and terminating in PSTN that can use IP networks and
protocols
• A call originating in a telephone network terminating in the Internet, and vice
versa
• Call handling/routing work in an all-IP environment .
PSTN Call Routing Using the Internet

Conceptual Requirement
VoIP Adapter Functionality

VoIP adapter would require the following:

• It has IP-stack based software and can generate IP packets as dictated by the upper
layer for transmission in Ethernet format.
• It has add-on software/functionality to generate Q.931 messages, voice
packetization, and the ability to generate certain sounds such as the dial tone that can
be heard through the phone receiver
 Addressing and Routing

Subscriber A connects a POTS telephone to the VoIP adapter.

• Subscriber A connects the VoIP adapter to the Internet.
• The VoIP adapter acquires an IP address, say, through the dynamic host configuration
protocol; we assume the address is 42.42.34.12.
• The VoIP adapter with IP address 42.42.34.12 starts an initial configuration session with
the gateway that has IP address 2.98.179.1. Since both these devices have Internet addresses,
any packet generated by the VoIP adapter will be routed through the Internet to destination IP. Through
this process, the telephone number
of the subscriber, +1-816-339-1706, is verified and stored in the VoIP adapter.
 Service Observations

If there is a loss of electricity at the home where the VoIP adapter is located, then the subscriber cannot avail
itself of the phone service; this is not to be confused with purposefully turning off the VoIP power much like
turning off a mobile phone. That is, always-available service is not completely possible unlike a plain
telephone connected to the central office switch where the electricity for the phone is directly supplied by the
central office.

• Subscriber A’s VoIP adapter is not tied to the IP address 44.44.34.12. The subscriber may
decide to take the VoIP adapter and connect to the Internet from anywhere in the world.
What gateway 2.98.179.1 really needs to be able to identify is that the VoIP adapter is authorized and that it
has an assigned E.164 number. This is possible since this information
is payload in a Q.931 message, which is then carried in an IP packet.

• The ability to take the VoIP adapter and connect to the Internet from anywhere is of great
interest to the subscriber. This way any local call to which the telephone number was
assigned is still a local call. In essence, this functionality provides location portability.

• The location portability, however, creates a problem in certain situations. For example,
Subscriber A who has the number +1-816-339-1706 takes the VoIP adapter anywhere and
connects to the Internet, and then dials for 911 emergency service. The 911 operator will
understand it as coming from the physical postal address on record. Recently, there have
been new discussions on how to to handle such issues; refer to [523] for recent developments.
Traffic Engineering

PER-CALL BANDWIDTH REQUIREMENT

ADDITIONAL FACTORS/OVERHEAD

CALL-CARRYING CAPACITY