CIT 5107: Internet of Things

Background
The following material from undergraduate curriculum necessary for the course:

ROUTING
In most structured networks, routing is not a significant problem given that the exact relationship between
the nodes is known and communication has to take a specific route. In Mesh topology and WANs, the
communication between nodes is not that obvious, making it necessary that some algorithm be used to
define communication between one node and another.

Routing Methods
Two methods are in use i.e
i) Deterministic/Fixed routing
ii) Adaptive routing
a) Deterministic/Fixed routing
Decisions on the route for onward transmission are made based on a routing table.
A packet enters the network with a destination address appended to it. Routing decisions are made
individually at each node by referencing the table, which indicates a preferred next destination based on
the network topology and other considerations. The table also includes alternative routes which will be
chosen in the event that the preferred route is congested or not available.

Destination address Data packet

2
B
5

A 1 3

A5 5 6
4
6 A2 5 1
A5 3 2 A6 6 5
A2 2 2
Packet address
A6 4 3
Alternative next node
Next node

1
In the diagram above, routing tables at two nodes have been given. Each table contains a single
alternative route in case the first one is not available. The routing tables take no account of the earlier
history of the packet, such as the number of nodes through which the packet has previously travelled. Its
only concern is with the forward information.
In this method, the message routes are generally unaffected by variations in network traffic. The routes
can be allocated centrally since they are never changed or changed infrequently.
Routes can be changed on a user log-in basis. Such a route is assigned depending on the network traffic
prevailing at the time and does not change for the duration of the users’ logged-in period.

b) Adaptive Routing
Adaptive routing has two options:
- Isolated adaptive routing
- Distributed adaptive routing
In adaptive routing, use is made of dynamic information on the traffic flow which is applied to modify the
fixed tables. The information is got from the traffic density measured at different exit ports on
neighboring nodes.
NB: It is difficult to make use of information from distant nodes in remote parts because of the problem
of transmitting the data over the network itself.
i) Isolated Adaptive Routing
Routing decisions are taken only on the basis of information available locally in each node consisting of:
 A pre-loaded routing table, as in fixed routing
 The current state of the on-going connection (i.e free or busy)
 The length of the packet queues awaiting the use of each node.
The routing algorithm is programmed to make a choice between alternative routes and expressed through
a calculation based on queue lengths and knowledge of the network topology giving a bias towards the
choice of better onward connections. This information is applied to modify the initial routing table held
within the node.

A node is considered to have available a primary and a secondary route to the next node. The number of
free spaces currently available on the output queue for the primary route is 2 and 4 for the secondary route
as in the diagram

2
With the weighting previously worked out from the routing table, the routing process will choose
between primary and secondary routes having a ratio of 3+ 2 and 1+4 1.e 5:5 giving an equal probability
of using either route.

E
E 3+2
Primary route
F
(Weight 3)
F
F

Either route taken

A A

Secondary route E

(Weight 1) E
E 1+4
F

E- Empty, F- Full

If the number of empty queue slots is, respectively 4 and 1, with the full ones remaining constant as
above, the probability ratio will be 3+4 and 1+1 giving 7:2 so that the bias will be towards the primary
route.
If one of the output queues is full, then the packet is sent to the other route irrespective of the number of
empty slots indicated. If both output queues are full, then the packet is discarded to be transmitted later
over the network. A message is in this case sent to the originating device for re-transmission.
ii) Distributed adaptive Routing
In distributed routing, the object is to find paths of least delay for the network traffic.

The average delay is measured at each node for every outgoing link to its neighboring nodes. This is
carried out every 10seconds and the information so obtained is propagated from the node to other nodes
in the system. Each node can use these estimates to compute its fastest path to ongoing nodes, in effect
provide a dynamic database of ‘shortest times’ for its local section of the network topology. These data
can then be used to obtain a decision on routing for all packets received, knowing the packet destination
addresses.

3
The delay incurred by the updating information packet circulated around the network can be considerable,
although independent of the total size of the network. In earlier versions of ‘Arpanet’, the update interval
was only 0.67 seconds and circulating and updating the delay table accounted for nearly 50% of the
bandwidth available for some lines. The interval between updating is now longer and in addition, delay
information is only transmitted when there is a significant change detected in the information to be
circulated from the last update table.

Flooding and random Routing

 Flooding does not use a formal systematic routing algorithm.
 These methods are used in communications where the configuration may not be known exactly or
is subject to modification at short notice.
With flooding, multiple copies of the packet are forwarded to all onward nodes from the node concerned-
except the one on which the packet was received. A previous calculation needs to be made on the shortest
number of nodes to be traversed to reach the destination – known as hop count. When this limit (or just
beyond it) is reached, the radiating packet is destroyed so that only the packet reaching the destination by
the shortest route is preserved.
I random routing, a random choice is made of an output path (other than the arrival path) along which to
direct the packet. This happens at the next node and so on. Again the hop count is used as a measure to
decide upon the elimination of a packet at a given node.

NB: Neither flooding nor random routing methods require information about the network
topology. In the case of flooding, the packet is guaranteed to reach the destination if this is
at all realizable.

Wireless (802.11) Network

Like its Ethernet and token ring siblings, 802.11 is designed for use in a limited geographical area
(homes, office buildings, campuses), and its primary challenge is to mediate access to a shared
communication medium—in this case, signals propagating through space.
It is designed to run over three different physical media—two based on spread spectrum radio and one
based on diffused infrared. The radio-based versions currently run at 11 Mbps, but may soon run at 54
Mbps. Spread spectrum is to spread the signal over a wider frequency band than normal, so as to
minimize the impact of interference from other devices.

4
Frequency spread uses two techniques:
- frequency hopping: involves transmitting the signal over a random sequence of frequencies; that
is, first transmitting at one frequency, then a second, then a third, and so on. The sequence of
frequencies is not truly random, but is instead computed algorithmically by a number generator.
The receiver uses the same algorithm as the sender—and initializes it accordingly—and hence is
able to hop frequencies in sync with the transmitter to correctly receive the frame.
- direct sequence, achieves the same effect by representing each bit in the frame by multiple bits in
the transmitted signal. For each bit the sender wants to transmit, it actually sends the exclusive-
OR (adding without carrying) of that bit and n random bits. As with frequency hopping, the
sequence of random bits is generated by a number generator known to both the sender and the
receiver. The transmitted values, known as an n-bit chipping code, spread the signal across a
frequency band that is n times wider than the frame would have otherwise required.

The third physical standard for 802.11 is based on infrared signals. The transmission is diffused, meaning
that the sender and receiver do not have to be aimed at each other and do not need a clear line of sight.
This technology has a range of up to about 10 m and is limited to the inside of buildings only.

A B C D

Consider the situation depicted in diagram, where each of four nodes is able to send and receive signals
that reach just the nodes to its immediate left and right. For example, B can exchange frames with A and
C but it cannot reach D, while C can reach B and D but not A. Suppose both A and C want to
communicate with B and so they each send it a frame. A and C are unaware of each other since their
signals do not carry that far. These two frames collide with each other at B. Unlike an Ethernet, neither A
nor C is aware of this collision. A and C are said to be hidden nodes with respect to each other.

5
A related problem, called the exposed node problem, occurs under the following circumstances. Suppose
B is sending to A in the diagram. Node C is aware of this communication because it hears B’s
transmission. It would be a mistake for C to conclude that it cannot transmit to anyone just because it can
hear B’s transmission, yet transmission to D would not collide with that to A.

802.11 addresses these two problems with an algorithm called Multiple Access with Collision Avoidance
(MACA). The idea is for the sender and receiver to exchange control frames with each other before the
sender actually transmits any data. This exchange informs all nearby nodes that a transmission is about to
begin. Specifically, the sender transmits a Request to Send (RTS) frame to the receiver; the RTS frame
includes a field that indicates how long the sender wants to hold the medium (i.e., it specifies the length of
the data frame to be transmitted). The receiver then replies with a Clear to Send (CTS) frame; this frame
echoes this length field back to the sender. Any node that sees the CTS frame knows that it is close to the
receiver, and therefore cannot transmit for the period of time it takes to send a frame of the specified
length. Any node that sees the RTS frame but not the CTS frame is not close enough to the receiver to
interfere with it, and so is free to transmit.

In completing the transmission, first, the receiver sends an ACK to the sender after successfully receiving
a frame. All nodes must wait for this ACK before trying to transmit. Second, should two or more nodes
detect an idle link and try to transmit an RTS frame at the same time, their RTS frames will collide with
each other. 802.11 does not support collision detection, but instead the senders realize the collision has
happened when they do not receive the CTS frame after a period of time, in which case they each wait a
random amount of time before trying again just as in Ethernet.