Journals & Books Register Sign in

Hierarchical Routing Related terms:

Dynamic Routing, Routing

Protocol, Communication
Networks, Triangulation, Reactive
Routing, Alternate Routing,
Telephone Network

View all Topics

Download as PDF Set alert About this page

Circuit-Switching: Hierarchical and Dynamic

Processing math: 100%
Call Routing
Deep Medhi, Karthik Ramasamy, in Network Routing (Second Edition),
2018

19.3.3 Call Control and Crankback

Hierarchical routing uses a progressive call control (PCC) mechanism. This
means that the call control is forwarded from one switch to another until
it reaches its destination unless the call cannot find any outgoing trunk at
an intermediate trunk; in this case, the call is lost. In other words, the
control of the call is not returned to the originating switch to try another
possible path.
Suppose we could return the control of a call from an intermediate switch
to the originating switch. This would mean that the network is providing
Originating Call Control (OCC); the functionality of returning a call to the
originating switch and trying another route is called crankback. With the
advent of dynamic call routing, the question of whether the network
should provide PCC or OCC and whether it should provide crankback also
arises.
Figure 19.6 illustrates how crankback works and its relation to OCC and
PCC. Consider a call arriving at switch 1 destined for switch 2. It can try
the direct link path 1–2. Suppose there is no bandwidth available on link
1–2 when the call arrives. The call will then attempt to use the next route
in the routing table 1–3–2. If link 1–3 has no available capacity, the call will
attempt the next route in the routing table 1–4–2; this overflow attempt is,
however, not a crankback. So what is a crankback? Consider a slightly
different situation. Suppose when the call attempted the second route 1–
3–2, it found bandwidth on the first link 1–3, and thus, the control of the
call is forwarded to node 3; however, on arriving at node 3 it was
discovered that there is no bandwidth available on link 3–2 for this call.
There are two possibilities: either send the control of the call back to the
originating switch 1 and let the originating switch decide what to do next
(for example, try another route such as 1–4–2), or drop the call. The
control of the call can be sent back to the originating switch 1 if the
network has originating call control, the process of reverting back to
switch 1 and trying another route is called crankback. If the network does
not have OCC, it must act as PCC. Thus, drop the call means that the call
on arriving at node 3 is lost due to non-availability of capacity on link 3–2;
this occurs due to PCC, the call control cannot be returned to switch 1. As
you will see later, some dynamic routing schemes provide OCC while
others do not.

Figure 19.6. Illustration of crankback.

View chapter Purchase book

Inheritance-based Adaptive Protocol for WSN

Information Routing
Youcef Touati, ... Boubaker Daachi, in Energy Management in Wireless
Sensor Networks, 2017

5.4 Critical analysis and conclusion

In this chapter, we have presented a hierarchical routing solution based on
information inheritance using an adaptive clusterization mechanism that
allows significant improvements to be made in minimizing energy
consumption, optimizing resources and therefore extending the lifetime
of the network. This approach is based on three main stages: a stage of
network deployment, initialization and recognition in which the sensor
nodes with well-defined features specific to the type of application are
deployed randomly in the operational environment, an adaptive
clusterization stage in which the different states of the sensor nodes are
taken into consideration, and a communication stage in which a single-
hop or multi-hop child–parent communication concept is proposed to
optimize the paths and the end-to-end timeframes for source–destination
packet conveyance.
In contrast to the standard LEACH protocol, where the nodes choose their
affiliated clusters several times, in the inheritance approach the node
chooses the CH with the maximum amount of energy for sending a single
message. The procedure for selecting new coordinator nodes does not
pass by the base station, but is carried out directly at the level of each
cluster. The former CH selects the new CH at the end of each round in
accordance with its energy level initially, and, if not, according to the
received signal strength when necessary.
Inheritance uses several metrics: the energy consumed, the network
payload, the end-to-end timeframes for conveying packets and the
strength of the signal from the RSSI indicator. The metric based on
calculating RSSI values is used not only in the recognition and
clusterization stages, but also in constructing optimal paths according the
child–parent concept. The advantage is that the RSSI technology is
widespread in the majority of WSN platforms, while also using relatively
low-cost energy.
However, in WSNs, communication links are asymmetrical and, given that
the strength of radio signals is measured upon reception, this can worsen
the performance of the proposed routing mechanism and cause errors in
estimating RSSI values. Those values can only be used to locate nodes
after a certain period of time as the stability of the signal is only belatedly
guaranteed. The time allocated for the recognition stage therefore
becomes important, as it accelerates energy consumption and thereby
minimizes the lifetime of the network.

View chapter Purchase book

Global and detailed routing

Huang-Yu Chen, Yao-Wen Chang, in Electronic Design Automation, 2009

12.5.2 Full-chip routing

Full-chip routing is typically a very complex combinatorial problem. To
make it manageable, many routing algorithms adopt a two-stage
technique of global routing followed by detailed routing. However, the
continuously increasing design complexity imposes severe challenges for
modern routers. The traditional flat framework does not scale well as the
design size increases. A modern chip may contain billions of transistors
and millions of nets. To cope with the scalability problem, routing
frameworks are evolving, and the hierarchical and multilevel frameworks
have become more and more popular for large-scale designs.
The hierarchical routing framework uses the divide-and-conquer approach
by transforming a large and complicated routing problem into a series of
smaller and simpler subproblems and then proceeds in a top-down,
bottom-up, or hybrid manner, which can be applied to both global and
detailed routing.
A top-down hierarchical global-routing framework has been proposed in
[Burstein 1983]. The algorithm recursively divides the routing regions into
successively smaller subregions, named super cells, and nets at each
hierarchical level are routed sequentially or concurrently and are refined in
the subsequent levels. Figure 12.21 illustrates an example of global
routing for a 3-pin net by the top-down hierarchical approach, in which
the routing region is recursively bisected into smaller super cells, and at
each level, the net is routed in terms of these super cells at that level. This
process is performed in a top-down manner until the sizes of super cells
reduce to that of global-routing tiles.

FIGURE 12.21. A level-by-level top-down hierarchical routing approach for a 3-pin net.

A bottom-up hierarchical routing method is developed in [Marek-

Sadowska 1984]. Initially, the routing region is partitioned into an array of
super cells. At each hierarchical level, the routing is restrained within each
super cell individually. When the routing at the current level is finished,
every four super cells are merged to form a new larger super cell at the
next higher level. This process continues until the top level containing the
whole chip is reached. Figure 12.22 shows the process of bottom-up
hierarchical routing for a 7-pin net, in which each solid rectangle
represents a super cell, and the 2*2 dotted subregions of the previous
level are merged together.

FIGURE 12.22. A level-by-level bottom-up hierarchical routing approach for a 7-pin net.

A major limitation in the top-down and the bottom-up hierarchical

approaches is that the routing decision made at one hierarchical level may
be suboptimal for subsequent levels. To alleviate this problem, Lin, Hsu,
and Tsai proposed a hybrid hierarchical approach that combines the
bounded maze-routing algorithm with both the top-down and bottom-up
hierarchical methods into a unified routing framework [Lin 1990]. Their
algorithm consists of three phases: (1) neighboring propagation, (2)
preference partitioning, and (3) bounded routing.
Phase 1 performs bounded maze routing by propagating W circles of
waves out of each pin, where W is a user-defined parameter. If the
connection is not found, Phase 2 recursively maps the pins and blockages
onto the adjacent upper level (see Figure 12.23a) and calls the bounded
maze-routing algorithm until a path is found. Then, the connected path is
mapped back to the lower level to preferred regions (see Figure 12.23b).
Phase 3 finds a routing path in the preferred regions (see Figure 12.23c).
Compared with pure top-down or bottom-up hierarchical routing, the
hybrid hierarchical approach has more global information to generate
better routing solutions.

FIGURE 12.23. An example of global routing by use of the hybrid hierarchical approach: (a)
Mapping pins and blockages up one level and then finding a routing path at the upper level.
(b) Mapping the connection at the upper level to the lower level to form the preferred
regions. (c) Finding a routing path in the preferred regions.

Although the hierarchical routing approach can scale to larger designs, it

has the significant drawbacks that the interactions among different
routing subregions are lacking and the routing decision at a level is
irreversible (i.e., cannot be refined at later stages), thus limiting the
solution quality. To remedy the deficiencies, researchers have proposed
the multilevel framework to handle large-scale routing problems. The
multilevel frameworks were first developed in [Cong 2001, 2002] for global
routing and in [Lin 2002] and [Chang 2004] for both global and detailed
routing. In the following, we introduce the routability-driven Λ-shaped
multilevel routing framework [Chang 2004].
The multilevel routing framework models the routing resource as a
multilevel-routing graph. At the beginning, the routing region is
partitioned into an array of rectangular subregions, each of which may
accommodate tens of routing tracks in each dimension (see Figure 12.24).
These subregions are called global cells (GCs). A node in the routing graph
represents a GC in the chip, whereas an edge denotes the boundary
between two adjacent GCs. Each edge is assigned a capacity according to
the physical area or the size of a GC. This routing graph is called the
multilevel-routing graph of level 0, denoted by G0, in which the subscript
represents the level.

FIGURE 12.24. The multilevel-routing graph.

The Λ-shaped multilevel routing framework consists of bottom-up

coarsening followed by top-down uncoarsening. The coarsening stage is a
bottom-up approach that iteratively groups a set of GCs in the multilevel-
routing graph. This process starts from the finest level (level 0) to the
coarsest level; at each level k, four adjacent GCk of Gk are merged into a
larger GCk+1 of Gk+1, and at the same time it performs resource estimation
for use at the k+1 level. Coarsening continues until the number of GCs at a
level is below a threshold. In contrast, the uncoarsening stage iteratively
ungroups a set of previously clustered GCs in a top-down manner. It
proceeds from the coarsest level to the finest level; at each level k, a GCk is
decomposed into four smaller GCk+1. Uncoarsening continues until the
finest level is reached. Figure 12.25 illustrates the Λ-shaped multilevel
framework.

FIGURE 12.25. The Λ-shaped multilevel routing framework.

Given a netlist, the multilevel routing first applies a minimum spanning

tree (MST) algorithm to decompose each net into 2-pin connections. At
each level k of the coarsening stage, global routing is first performed for
the local 2-pin connections (those connections that entirely sit inside a
GCk), and then the detailed router is used to determine the exact wiring.
Let the multilevel-routing graph of level 0 be G0 = (V0, E0), and the global-
routing result for a local connection be Re = {e ∈ E0|e is the edge chosen
for routing}. For the congestion control, the cost function α : E0 → R is
applied to guide the routing:
α(R e) = ∑ e∈R c e (12.4)
e

where ce is the congestion of edge e and is defined by

c e = 1 / 2 (p e-d e) (12.5)

where pe and de are the capacity and density associated with e, respectively.
Note that we always search the shortest global-routing path between two
pins in the coarsening stage therefore (i.e., monotonic routes or no
detours); therefore, the wirelength is the minimum, and thus the
wirelength is not included in the cost function at the global routing stage.
This cost function can guide the global router to select a path with smaller
congestion.
After the global routing is completed, the detailed routing applies a
simultaneous pathlength and via minimization (SPVM) algorithm to perform
modified maze routing that simultaneously considers the pathlength and
via minimization. For better circuit performance, it is desirable to
minimize the number of vias used in a routing path, because vias typically
have significantly larger RC delay than metal wires. The SPVM algorithm
can find a shortest path with the minimum number of bends/vias, if such
a path exists. It associates each basic detailed routing region u (could be a
grid cell in grid-based routing or a basic routing region defined by the
wire pitch in gridless routing) with two labels d(u) and b(u), where d(u) is
the distance of the shortest path from the source s to u, and b(u) is the
minimum number of bends/vias along the shortest path from s to u.
Initially, d(s), b(s) = 0, and d(u), b(u) = ∞, ∀ u ≠ s. In the filling phase of
maze routing, the computation of label d is the same as the original maze-
routing algorithm. Let u be a basic routing region on the wavefront of
wave propagation and v a neighboring basic-routing region of u. The
predecessor routing region of u is the region from which the wavefront
was propagated for obtaining the minimum b(u). The propagation
direction of u is the direction from the predecessor routing region of u to
u. The computation of b(v) is shown in Algorithm 12.2.

Algorithm 12.2
Computation of b(v) in the SPVM Algorithm

The basic idea is to compare the distance label d first and then compare
the bend/via number label b. The value b(v) of a neighboring routing
region v with d(v) < d(u) remains unchanged, because the path from s
through u to v is not the shortest path between s and v. The retracing
phase is the same as that of the original maze-routing algorithm. Note
that there may be several shortest paths with different numbers of
bends/vias. The wave-propagation phase always keeps track of the shortest
path with the minimum bend/via number to allow the retracing phase to
find such a path.
When the global and detailed routing is performed at level k, four adjacent
GCk are merged into a larger GCk+1 and at the same time resource
estimation is performed for use at the next level k + 1. Because the global
routing, detailed routing, and resource estimation are integrated together
at each level, the routing resource estimation is more accurate, thus
facilitating the solution refinement (e.g., the rip-up and reroute processes)
at the uncoarsening stage. Algorithm 12.3 gives the algorithm of the Λ-
shaped multilevel routing framework [Chang 2004].

Algorithm 12.3
Λ-Shaped Multilevel Routing Algorithm

View chapter Purchase book

Wearable Wireless Sensor Networks for

Emergency Response in Public Safety
Networks
Muhammad Mahtab Alam, ... Elyes Ben Hamida, in Wireless Public Safety
Networks 2, 2016

3.4.3.2 BBN routing protocols

In higher layers, routing in PSN has to handle frequent node
disconnections and reconnections, which will influence the capabilities
required for the routing protocols. This should be ensured without
causing excessive traffic overhead or computational burden on the power
constrained devices. To meet the requirements detailed in [IEE 12], and
referring to the effective networking model presented in [ALA 14], a variety
of ad hoc networks are compared in [REI 14]. Mobile Ad hoc Networks
(MANETs) are the classical approach regarding the implementation of
public safety networks, broadcasting communications with multi-hop
communications based on reactive and proactive routing protocols and
consequently covers the body-to-body (BBN) routing aspects. A recent
study on the evaluation of MANETs in emergency and rescue scenarios is
investigated in [QUI 13]. The assessment of MANETs routing protocols
refer to the proactive, reactive, hybrid and hierarchical routing protocols
classes. Further on, we will discuss whether this class of the routing
protocol is appropriate to the PSN applications context or not.
Proactive routing protocols, such as Optimized Link State Routing
protocol (OLSR) and Destination Sequenced Distance Vector (DSDV),
continuously exchange information to keep up-to-date routes of all
network nodes, and is important in the case of victims’ evacuation or
rescue missions. However, this may negatively affect the bandwidth
utilization and significantly decrease the batteries lifetime especially in
small devices. On the other hand, reactive routing protocols such as Ad
hoc On-demand Distance Vector (AODV) and Dynamic Source Routing
(DSR) are characterized by their two mechanism components: route
discovery and route maintenance. The route discovery process of the latest
version of AODV (i.e. version 2), called also DYMO, is illustrated in Figure
3.8. The latency to initiate the communication and the delay to detect
network changes excludes, reactive routing protocols from tactical and
rescue context. Moreover, these flat routing protocols are not appropriate
to critical and rescue context. There are also hybrid routing protocols
which combine two or more proactive and reactive routing protocols.
Clustering or hierarchical routing protocols is another class which divide
the network into zones or clusters, and then dedicate some important
tasks of the network to a special node (called Cluster Head) to reduce local
calculations and overhead for the classic nodes. The last two categories are
instantiated from the reactive and proactive approaches in order to reduce
the power consumption and increase network performance, with
consideration to the high mobility of the nodes. Cluster Based Routing
Protocol (CBRP) is a considerable hierarchical routing protocols.

Figure 3.8. Route Discovery Process (AODV-v2)

For public safety networks, an ad hoc cognitive radio (CR) based spectrum-
aware routing protocol is proposed in [BOU 12]. This routing protocol
comes up with the use of the white spaces spectrum resources in TVWS
(Television White Space). It presents a specific adoption of AODV routing
protocol due to the unpredictable availability of the TVWS which requires
a hop-by-hop routing. The contribution lies in the fact that during the
route request (RREQ), the proposed routing protocol includes the TVWS
availability of nodes. Each RREQ will inform about the source nodes and
TVWS availability. It evaluates the performance in a specific scenario, and
validates the adoption of AODV under the controlled simulation
conditions in [BOU 12].
Delay Tolerant Network (DTN) is a network based on nodes encountering
where a node waits until it encounters another node to deliver the packets.
Characterized by its latency, DTN is suitable for low density networks [REI
14], but not in critical and rescue missions. All forwarding mechanisms of
DTN are based on opportunistic communications where routes are built
dynamically through any encountered nodes (e.g. Epidemic forwarding,
PRoPHET forwarding, MaxPropand TTR) [MAR 13].
In the “statement of requirements” of the PSC, as detailed in the
SAFECOM, various operating conditions and challenges are highlighted
which are exploited by the authors in [BOH 09], to propose a QoS
enhancement for ad hoc routing in the rural PSN. QoS-Aware source
routing (QASR) protocol is based on dynamic source routing (DSR)
because of the integration of the entire path in the packet. The integration
of the bandwidth and delay estimation algorithms is realized in the route
discovery process. QASR then selects a path for the route reply (RRep)
from the available discovered paths. QASR estimates the available
bandwidth and delay from the information collected from all nodes.
Moreover, since the GPS location systems are quite possibly implemented
in the most mobile nodes (in used public safety operations), the distance
between a node and its neighbors is known. Thus, the received message
will be rebroadcasted if the calculated distance (between the given node
and its neighbors) is higher than the interference range, otherwise the
message will be dropped. In case of an unknown node position, there will
be a random generated number compared with the threshold to decide
whether the message will be rebroadcasted or not. Subsequently, three
parameters will define the route selection: estimated available bandwidth,
estimated delay and the node speed. The comparative analysis on
abovementioned protocols is highlighted in Table 3.3.

Table 3.3. Considered protocols for BBN and comparison

Parameter/Metrics Routing Protocols

AODV OLSR QASR CBRP

Routing class Reactive Proactive Routing Dynamic Source Hierarchical Routing

Routing Routing

Route discovering On- initiated with network initiated with

Intra-cluster: initiated with
demand initialization network
network initialization
initialization
Inter-cluster: On-demand

Control message RREQ, Hello, TC Hello, TC Hello, RREQ, RREP, RERR

RREP,
RERR

Link support Symmetric Symmetric or Symmetric or Asymmetric

With MPR:
Asymmetric
Symmetric
With others:
Symmetric or
Asymmetric

Central No No No Cluster head

administration

Require sequence Yes May be used but not Yes Yes

number mandatory

Performance Shortest Cost Bandwidth, Delay Shortest path

criteria path and Speed

Energy Aware Routing in Heterogeneous Multi-Hop Public Safety

Wireless Networks is a recently proposed routing protocol that considers a
compromise between the energy consumption and the routing delay as
the routes selection criteria. Authors of the proposed routing protocol
present a strategy for route path selection optimization [FED 14]. This
optimization includes the selection of the minimum energy link cost route
path and the maximum network lifetime. An Energy Aware Routing
technique based on an on-demand routing protocol in this protocol
includes the energy related metrics (i.e. remaining battery level, energy
cost per bit, etc.). Indeed, a destination node that receives multiple route
requests from the same source node selects the best route based on the
optimization strategy for route path and unicasts the route reply through
that best route. A hop-limit is considered for the route request to avoid
discovery packets flooding. This hop-limit depends on the network size
and density. The proposed Energy Aware Routing technique also allows
the destination of the route request packet to explore the various networks
technologies encountered by the route request while being broadcasted.
Such collected information on the network gives the nodes an overall view
to be considered while the routes computation process and the route reply
message unicast.
In future the routing protocols are required to consider latency, reliability,
mobility, thermal-effects and energy consumption as collective metrics. In
this context, recently an interesting study was conducted in which various
classes of routing approaches (including ad hoc MANET, geographical
based routing and data centric based routing) were compared [ARB 15a].
In addition, various wireless technologies including IEEE 802.15.4 WSN,
IEEE 802.11 WiFi, IEEE 802.15.6 WBAN are utilized and compared for
analysis of the multiple performance metrics. Similarly, in [ARB 15b],
authors investigated the data dissemination problem in the BBN context
and compared clustered-based techniques versus fully distributed
techniques. To summarize, the emerging challenges in heterogeneous
networks lead to the need for a cross-layer networking solution which will
be in charge of selecting an appropriate communication technology in
PSN.

View chapter Purchase book

Networking Support
Dan C. Marinescu, in Cloud Computing, 2013

7.14 History notes

The Internet is a global network based on the Internet Protocol Suite
(TCP/IP); its origins can be traced back to 1965, when Ivan Sutherland, the
head of the Information Processing Technology Office (IPTO) at the
Advanced Research Projects Agency (ARPA), encouraged Lawrence
Roberts, who had worked previously at MIT’s Lincoln Laboratories, to
become the chief scientist at ISTO Technologies and to initiate a
networking project based on packet switching rather than circuit
switching.
In the early 1960s Leonard Kleinrock at the University of California at Los
Angeles (UCLA) developed the theoretical foundations of packet networks
and, in the early 1970s, for hierarchical routing in packet-switching
networks. Kleinrock published the first paper on packet-switching theory
in 1961 and the first book on the subject in 1964.
In August 1968 DARPA released a request for quotation (RFQ) for the
development of packet switches called interface message processors
(IMPs). A group from Bolt Beranek and Newman (BBN) won the contract.
Several researchers and their teams including Robert Kahn from BBN,
Lawrence Roberts from DARPA, Howard Frank from Network Analysis
Corporation, and Leonard Kleinrock from UCLA, played a major role in the
overall ARPANET architectural design. The idea of open-architecture
networking was first introduced by Kahn in 1972, and his collaboration
with Vint Cerf from Stanford led to the design of TCP/IP. Three groups,
one at Stanford, one at BBN, and one at UCLA, won the DARPA contract
to implement TCP/IP.
In 1969 BBN installed the first IMP at UCLA. The first two nodes the
ARPANET interconnected were the Network Measurement Center at
UCLA’s School of Engineering and Applied Science and SRI International
in Menlo Park, California. Two more nodes were added at UC Santa
Barbara and the University of Utah. By the end of 1971 there were 15 sites
interconnected by ARPANET.
Ethernet technology, developed by Bob Metcalfe at Xerox PARC in 1973,
and other local area network technologies, such as token-passing rings,
allowed the personal computers and workstations to be connected to the
Internet in the 1980s. As the number of Internet hosts increased, it was no
longer feasible to have a single table of all hosts and their addresses. The
Domain Name System (DNS) was invented by Paul Mockapetris of
USC/ISI. The DNS permitted a scalable distributed mechanism for
resolving hierarchical host names into an Internet address.
UC Berkeley, with support from DARPA, rewrote the TCP/IP code
developed at BBN and incorporated it into the Unix BSD system. In 1985
Dennis Jennings started the NSFNET program at NSF to support the
general research and academic communities.

View chapter Purchase book

Call Routing in GSTN

Deep Medhi, Karthik Ramasamy, in Network Routing (Second Edition),
2018

25.7.2 Illustration of Call Routing

We will illustrate the above concept through the use of the North
American numbering plan, dialing from a residential landline phone.
While this scenario is not applicable in practice anymore in the United
States, we use it purely for illustration.
First, it is important to note that this is a conceptual logical view of the
steps taking place inside a TDM switch while noting that different switch
vendors may implement the entire call processing steps/logics somewhat
differently.
Assume that a call originates from the number +1-816-328-9999. The user
dials basic categories of calls such as local calls 235-2006, 913-235-2006,
long-distance calls 1-351-235-2006, and international calls 011-91-11-
623-45678, in addition, to special service calls 911, and toll-free calls – 1-
800-212-XXXX.
The network architecture as viewed from the originating TDM switch, and
taking into account SS7 networking and point code addressing, is
presented in Figure 25.12. For simplicity, just one signal transfer point
(STP) is shown to be connected to each SSP; that is, the mated STP is not
shown. Assume that the point code for the originating SSP is 251.1.3.
When the user (i.e., +1-816-328-9999) dials a telephone number for
processing at the originating TDM switch (PC 251.1.3), it must follow the
routing rule for call routing that is based on hierarchical routing; that is,
first do direct trunkgroup routing if there is a high usage trunkgroup and
then overflow to alternate routing. For a local call, the final trunkgroup is
the trunkgroup to the local tandem switch for alternate routing; for calls
outside its region, the final trunkgroup is the trunkgroup to an
appropriate toll switch. For convenience, this view of hierarchical routing
is shown in Figure 25.14 using the routing rule and node notations
discussed earlier in Section 19.2 in Chapter 19. To mimic this behavior in
an SS7-enabled environment, the originating TDM switch maintains a
lookup table for leading digits dialed to the immediate next hop SSP,
along with an alternate SSP for overflow routing. In our example, SSP
251.1.4 is a local tandem switch for overflow routing for local traffic. In
Figure 25.13, we present the example tables maintained at the originating
TDM switch for call processing; their usage will be discussed below
through a set of scenarios.

Figure 25.12. Call and message routing for calls dialed from 816-328-9999.

Figure 25.13. Call routing decision: Conceptual view through table lookup.

Figure 25.14. Hierarchical routing view at the TDM switch level.

Dialing 235-2006
Consider first that the user dials 235-2006. Then, the originating TDM
switch identifies the next hop SSP to be 251.1.2 using Figure 25.13(a); it
then uses Figure 25.13(b) to identify the trunkgroup ID as 1 and checks if
an idle circuit is available. We assume that there is a circuit available so
that direct routing can be performed. The TCIC code for this circuit is then
generated; the IAM message originating at SSP 251.1.3 includes this
TCIC; in addition, this IAM message contains considerable information,
most critical of which is that it includes the called number 816-235-2006.
The prepared IAM message is sent to SSP 251.1.2. From Figure 25.12,
note that these two SSPs are shown to be connected by an F-link in the
SS7 network; thus, this IAM message is routed on this link. On receiving
this message at SSP 251.1.2, it checks the content of the message and
determines that this is the termination TDM switch for this call.
Dialing 913-235-2006
Now consider dialing 913-235-2006, a local number within geographic
proximity of the originating number. First, assume that the call would be
directly connected like the call to 816-235-2006. In this case, the
destination SSP is identified as 251.2.6, and thus, the originating TDM
switch will identify the trunk group identifier as 3 and check if there are
circuits available on trunkgroup identifier 3. One difference, however, is
that there is no F-link in the SS7 network between SSP 251.1.3 and SSP
252.2.6. Thus, the IAM message on the SS7 network would be forwarded
to the STP 251.1.0, for further forwarding to SSP 252.2.6.
Consider the next possible scenario for the call to 913-235-2006, i.e., if
there are no circuits available on the high usage, direct trunkgroup
(identifier 3) from the originating TDM switch to the destination TDM
switch. Based on the routing rule and lookup (see Figure 25.13), the
originating TDM switch would try the trunkgroup to the tandem switch
with PC 251.1.4 to see if there are any available circuits. We assume that a
circuit is available on the trunkgroup to the tandem switch. Then, due to
the progressive call control (PCC) functionality, the control of the call
would need to be forwarded to SSP 251.1.4. This can be accomplished
using the IAM pass along method, i.e., the IAM message is sent to SSP
251.1.4 indicating in its content that the called number is 913-235-2006
with originating point code (OPC) as 251.1.3 and destination point code
(DPC) as 251.1.4. On receiving this IAM message, SSP 251.1.4 would
check if it is meant for itself; based on its own routing table, it would
know that the call set up message must be forwarded to the final
destination, SSP 252.2.6. Thus, SSP 251.1.4 will create a new IAM message
indicating itself as the OPC and setting the DPC as 252.2.6; this message
will contain the newly determined TCIC code for the available circuit on
the trunkgroup between SSP 252.1.4 and SSP 252.2.6 and will also contain
the same called number 913-235-2006. Furthermore, SSP 251.1.4 will
maintain in its memory a connectivity/mapping state table between the
IAM message received from SSP 251.1.3 and the one sent to SSP 252.2.6.
When SSP 252.2.6 receives the IAM message from SSP 251.1.4, it
determines that it is indeed terminating at this node, and it then sends an
indication to telephone 913-235-2006 about ringing, and returns an ACM
message to 251.1.4 with OPC 252.2.6 and DPC 251.1.4. When SSP 251.1.4
receives this ACM message, it checks its connectivity/mapping state table;
thus, SSP 251.1.4 will in turn generate an ACM message with OPC
251.1.4. and DPC as 251.1.3. When the originating TDM switch 251.1.3
receives this ACM message, it knows that the call is connected.
From the above discussion, we see that the IAM/ACM message content in
regard to OPC and DPC changes at the intermediate TDM switch; this is
essential due to hierarchical call routing with progress call control and to
avoid any confusion at the destination TDM switch. For instance, suppose
that SSP 251.1.4 did not change the OPC field but changed only the DPC
field to 252.2.6 in the IAM message sent to 251.1.6. Then SSP 252.2.6, on
receiving this message, would see the OPC as 251.1.3 and would think
that the TCIC value is on the trunkgroup between the originating SSP
251.1.3 and itself, which is certainly not the case.
Dialing Emergency Service Number 911
Next, we consider dialing 911. In this case, we assume that the 911
operator trunks are connected off the local tandem switch with PC
251.1.4. Thus, in this case, the IAM message would need to be routed
through STP 251.1.0 to SSP 251.1.4 since there is no direct F-link between
SSP 251.1.3 and SSP 251.1.4. Once the IAM message arrives at SSP
251.1.4, it determines that the call is to be put on the trunkgroup to the
911 operator center, known as the Public Service Access Point (PSAP).
Dialing 1-351-235-2006 and 011-...
Next we consider dialing a long-distance or an international number. By
inspecting Figure 25.13, we see that for both cases the toll switch with SSP
251.3.8 is the next TDM switch, and the call will be attempted on the final
trunkgroup to this switch (there is no high usage trunkgroup in this case).
The originating switch (SSP) will create an IAM message with OPC as
251.1.3 and DPC as 251.3.8. Since there is no F-link between these to
SSPs, this IAM will be forwarded to STP 251.1.0, for further forwarding to
251.3.8.
Dialing 1-800 Number
Finally, consider the case where the user dials a 1-800 number. From
Figure 25.13(a), the originating switch would know that the next hop SSP
is 251.1.7. By consulting Figure 25.13(b), the switch would know that it is
required to generate an SCCP message to SCP node 251.1.7 to obtain title
translation for the 1-800 number. This SCCP message will be forwarded to
STP 251.1.0, which will in turn be forwarded it to PC 251.1.7 (see Figure
25.12). In the response message to SSP 251.1.3, SCP 251.1.7 will indicate
that the mapping for this 1-800 number is 1-557-433-XXXX. On receiving
this message, the originating TDM switch again consults Figure 25.13(a)
and finds out that the SSP is 252.3.8 (same as the number that starts with
+1-351). It will then handle the call in the same way as it would handle a to
call to 1-351-235-2005 as described earlier.
Today's 800 number handling is much more sophisticated than what is
described so far. This will be discussed later in Section 25.10.

View chapter Purchase book

Routing Information for Energy Management

in WSNs
Youcef Touati, ... Boubaker Daachi, in Energy Management in Wireless
Sensor Networks, 2017

3.3 Critical analysis

In this chapter, we have concisely illustrated the issues and challenges in
WSNs, before exploring routing techniques developed in recent years by
drawing attention to factors improving and/or worsening network
performances and functioning.
Increasing the performance of a WSN in terms of longevity, connectivity
and robustness requires that essential factors such as the energy
consumed and the calculation and memory resources be taken into
consideration. It is therefore necessary to find alternatives, particularly
software alternatives, as once a network is deployed it cannot withstand
repeated human interventions for each malfunction it might have, i.e.
battery failure, environmental influence and so on. Moreover, these
malfunctions can cause significant topological changes, requiring a
complete reorganization of the network in order to ensure the exchange
and conveyance of data. Proposing suitable solutions based on routing
can bring about important advances in improving the rate of information
exchanges by considering topological changes and stabilizing the network
operation over time by making optimal use of energy consumption. A
data-centered routing intended primarily for networks with flat topologies
and which are known for their high tolerance of breakdowns during route
construction, with very low maintenance costs, can be considered as a
solution. Nonetheless, it uses a significant number of control messages to
ensure that the network functions well without actually improving
scalability and therefore degrading performance in some cases. Moreover,
the nodes located close to the base station are often required and if their
loss is accelerated it can lead to an overconsumption of energy, reducing
the lifetime of the network. On the other hand, the hierarchical routing
known for the advantages it offers in scalability and transmission
efficiency can provide solutions in optimizing energy consumption,
particularly for dense heterogeneous networks. During the deployment
stage, CHs, which are more powerful in terms of energy, bandwidth and
memory and which function differently to ordinary nodes, are formed and
then chosen. They allow the number of messages conveyed to the target
to be optimized through the aggregation of data from different sources,
considerably reducing energy consumption. Location-based routing
identifies the location of nodes in the network by estimating the distances
between neighboring nodes, either by exchanging RSSI information and
using empirical models made up of environmental and energy
parameters, or directly using a satellite system and, in this case, the
sensors embed low power GPS receivers.
Furthermore, the implementation of a routing protocol must take into
account and, if possible, optimize all sources of energy consumption,
which are numerous and do not only concern the transmission and
reception of data [HAL 09]. These sources include processing carried out
by the processor, the detection of environmental information, transitions
between the different states of a sensor such as the standby and active
modes, reading and writing memory, actuators and, of course, the
creation of clusters in the case of hierarchical architectures. Each of these
sources consumes a different amount of energy, and the percentages of
energy loss for a WSN with 100 nodes shown in Figure 3.8 are proof of
this.

Figure 3.8. Energy consumption rate in a WSN. For a color version of this figure, see
www.iste.co.uk/touati/energymanagement.zip

Over 50% of energy losses are in fact caused by the radio; that is, the
transmission and receipt of data. The greater the distances, the more
significant the amount of energy required to route information.
Depending on the WSN applications, the energy sources can be taken into
consideration or not. For instance, an application for monitoring forest
fires does not consider energy used on actuators, or even on storing or
saving data, and, in this case, over 70% of the energy used goes on radio
transmission.
Maximizing and estimating the lifetime of a WSN require the
implementation of an energy consumption model that suits the
application in question. Several pieces of research have examined this
context and proposed more or less specific models that specify certain
sources of energy consumption. The Heinzelman model [HEI 02], for
example, does not take into account energy models related to processing
or radio transmission. Miller [MIL 05] considers the status transitions of
sensor nodes. The Zhu model [ZHU 03] incorporates energy spent on
detection in the energy used on processing and communications. Models
related to detection, memory and actuators are rarely considered.
Proposing a routing solution must therefore occur primarily in accordance
with the context and the type of application considered by identifying the
sources of energy consumption and, above all, the parameters to be taken
into account. It is thus important to not choose a specific routing
procedure at the expense of another, but to draw from the entire range of
existing solutions.
In the chapter to follow, we will look at dense networks, particularly those
structured hierarchically, and with the objective of improving performance
in energy consumption and information routing delays on the basis of
consumption models related to processing, radio and cluster formation.

View chapter Purchase book

About ScienceDirect Remote access Shopping cart Advertise Contact and support Terms and conditions Privacy policy

We use cookies to help provide and enhance our service and tailor content and ads. By continuing you agree to the use of cookies.
Copyright © 2020 Elsevier B.V. or its licensors or contributors. ScienceDirect ® is a registered trademark of Elsevier B.V.