Advertisment

Having WILL Power

author-image
Voice&Data Bureau
New Update

In the developing countries, the
teledensity is far below the figure of 20 percent recommended by ITU. These countries
include much of South America and Africa, parts of the Middle East, India, and remote
parts of Russia and China. These countries have limited teledensity because of lack of
financial resources for building up a large wired network. Moreover, in such countries,
people have restricted their interaction and needs within their own groups and
communities, thus limiting their social and economic growth.

Advertisment

alt="https://img-cdn.thepublive.com/filters:format(webp)/vnd/media/post_attachments/52209add08b427e3371501d3378c4351fbafe1647b9fb8dda52e6f07500c1c98.gif (40705 bytes)" align="right" hspace="6" vspace="6"> face="Times New Roman">Under these circumstances, the key requirements of a network are
low capital and easy and faster implementation. Furthermore, difficult terrain and
climatic conditions also impose restrictions on the implementation of wired networks. For
such countries, high bandwidth and multi-port lines per house are not in much demand. The
key issues for all such countries in this category are faster and low-cost implementation
of telecommunication infrastructure expecting economic growth, with the requirements
ranging from simple voice only capability to high-speed Internet and multimedia
requirements.

Indian telecom scenario is a typical case
of this category. In India, average teledensity is less than 2 percent. Rural teledensity
is less than 0.5 percent. This is with the factor that around 50 percent of the Indian
villages are without a telephone number.

The New Challenges COLOR="#000000">

Advertisment

Copper has provided access to the
telephone network for the past 100 years. Though sufficient changes have taken place in
the telecom scenario, the copper local loop has remained largely unaffected. Though
radio-based access is not new and is in use since long, the local loop distribution from
the radio terminal to a house is still based on copper cable.

This is where Wireless In Local Loop
(WILL) is replacing copper, because of low-cost, fast and easier, deployment, and greater
ease of reach to difficult geographic terrain. DoT in India has already taken a decision
to introduce WILL-based systems. It is estimated that over the next five years, the
domestic WILL market will have between two million and 2.5 million subscribers. This is
based on the expectations that 20-25 percent of all new lines being rolled out will use
WILL technologies.

bgcolor="#000000">Codecs–A
Comparison


face="Arial" SIZE="1" color="#FFFFFF">Codec face="Arial" SIZE="1">PCM face="Arial" SIZE="1">ADPCM face="Arial" SIZE="1">GSM full rate face="Arial" SIZE="1">Enhanced GSM full rate face="Arial" SIZE="1">GSM half rate face="Arial" SIZE="1">TETRA
color="#FFFFFF">Data Rate (kbps) color="#FFFFFF">Quality color="#FFFFFF">Comments
color="#000000" size="1">64 Excellent SIZE="5" COLOR="#ffffff">
color="#000000" size="1">32 Virtually
PCM
WILL
applications
13 SIZE="5" COLOR="#ffffff"> Noticeably
worse than PCM
13 SIZE="5" COLOR="#ffffff"> Better
than GSM full rate
To
be introduced soon
6.5 SIZE="5" COLOR="#ffffff"> Slightly
worse than GSM full rate
Unlikely
to be widely used
4.4 SIZE="5" COLOR="#ffffff"> To
be demonstrated
Advertisment

However, initially WILL had to prove
itself as a viable alternative to the existing copper. The quality of service in terms of
traffic handling capacity as well as voice quality needed to be equivalent to that in a
copper loop. The key requirements are:

  • Plain old telephony,
    with voice quality as good as wire-line system

  • N-ISDN, with data rates
    of 144 kbps providing at least two lines, either of which can be used for voice or data.
    SIZE="1">

  • Fax transmission SIZE="1">

  • Supplementary
    services–call divert, follow me, call waiting, barring of incoming and outgoing
    calls, etc.

  • Internet service SIZE="1">

  • Leased line (supporting analog
    as well as digital standards).

Advertisment

Separate links and circuits have to be
provided for high-speed Internet access and leased lines. For voice telephony in local
loop, toll quality voice is required. For obtaining this either 64 kbps PCM or 32 kbps
ADPCM should be used. Traffic generation per telephone subscriber should be assumed from
0.05 erlang to 0.15 erlang with call blocking rate less than 1 percent.

face="Arial" COLOR="#ffffff" size="4">The Economics of a WILL System–A
Comparison
SIZE="2" COLOR="#000000">Wired System

face="Arial">For providing wired access for telephony, it is mandatory to provide
telephone cable from distribution point to each house. For doing this, the perquisites are
digging of roads, laying of cables, mounting of poles, repairing of the damaged roads,
etc. In the densely populated city area, cable laying is much more expensive and labour
intensive.

As a general guideline, it can be assumed that laying of a cable in
city area costs around $30 per metre. The cost to connect a house is given by its distance
from the last house connected, multiplied by the cost per metre of laying the cable. As a
typical case, assume that the houses in a particular area are 10 metres apart (end-to-end)
with 5 metre long front courtyards. It can be assumed that digging of road costs $30 per
metre, whereas inside the house it costs $20 per metre. The termination cost for each
house can be approximately be $40-50. In case of 100 percent penetration, cost per home
would be $450 (10 m x $30 + 5 m x $20 + $50 = $450).

If the penetration drops to 20 percent, it is necessary to pass four
other houses before installing the cable in a house. Thus the laying cost per house
becomes $1,650 (5 x 10 m x $30 + 5 m x $20 + $50 = $1,650).

Moreover, maintenance cost of wired system is also very high. The
cables suffer from flood or get damaged due to digging of roads for various other
purposes. The maintenance cost is estimated around 5 percent of the installation cost per
year.

face="Arial">Wireless System

In case of a wireless system,
the operator erects a transmitter. For providing coverage to subscribers, radio units are
mounted on the side of houses.

Typically, transmitter installation costs around $150,000
(conversion of landed cost) and the subscriber units cost around $400 each. Let us assume
there are 1,000 houses in the coverage area of the transmitter (medium population
density). With 100 percent penetration, the transmitter cost per house comes to be around
$150 each. Hence, per subscriber cost becomes $150 + $400 = $550.

However, in actual practice, 100 percent penetration is too
optimistic. Assuming the penetration of 20 percent, transmitter cost per subscriber
becomes $750. Thus total cost per subscriber becomes $1,150, which is much less than
$1,650 in case of the wired line system.

Figure 1 shows the advantages of a WILL system as the penetration
falls in case of medium housing density.

However, in case of a densely populated area or in large buildings
with lot of subscribers/offices residing in, the cost of connecting houses close to
switch/transmitter will be lower in case of copper cable rather than radio for individual
subscriber. Practically, speaking in a radius of around 250 metre, with medium to large
density, cable will be better solution.

This leads to hybrid environment, which may become the most
cost-effective and practical solution for providing telephony.

The system should support high subscriber
density ranging from 1,000 subscribers per sq. km to 100,00 subscriber per sq. km. For
supporting all these factors, bandwidth availability is the key issue.
SIZE="4" COLOR="#016077">

Advertisment

Bandwidth Requirements and Micro Cells SIZE="2" COLOR="#000000">

Bandwidth requirements depend mainly on
the number of subscribers to be served and traffic generated.
src="infrast2.gif" width="314" height="223" alt="infrast2.gif (10240 bytes)" align="left"
hspace="6" vspace="6">

Assuming the use of 32 kbps ADPCM voice
coding, 30 kHz of frequency spectrum will be required for each way. Thus for two-way
communication two channels will be required raising the bandwidth requirement to 60 kHz.
Assuming that with CDMA or DCA, TDMA (1/1) frequency reuse pattern is adapted. Other
assumption includes 1,000 subscribers per sq. km., with 0.1 erlang traffic. Let us say
that we wish to serve subscribers within a radius of 4 km. The total area served will be
around 50 sq. km. and the total number of subscribers served by that base station will be
50,000 (50 x 1,000 subscribers per sq. km.). With 0.1 erlang traffic per subscriber the
number of channels required per cell is approximately 5,000 (50,000 x 0.1). The total
bandwidth required per cell will be 300 MHz (5,000 x 60 kHz) per cell. Moreover, as the
subscriber density increases, the bandwidth requirements will also increase.

Advertisment

It can be seen that micro cells help to
serve larger subscriber density. Assume that cell radius has been reduced to 500 m. For a
subscriber density of 1,000 subscriber per sq. km., the number of subscribers served per
base station will be approximately 800 subscribers.

With 0.1 erlang traffic per subscriber,
the bandwidth required will be approximately 4.8 MHz per cell (800 x 0.1 x 60) only.

Thus, we can see that micro cells are the
solution for serving large number of subscribers. But micro cell architecture has its own
problems like inter-cell interference and restrictions on frequency re-use. However, these
problems can be overcome by proper cell planning and location of base station.
SIZE="4" COLOR="#016077">

Advertisment

TDMA or CDMA

Since the beginning of the Nineties, there
is a debate going on about whether CDMA or TDMA is the best access technique. There are
various factors which are to be considered for these two techniques, viz. capacity
comparison, range, sectorization, frequency planning, and micro cell structure.
SIZE="2">

Capacity Comparison

The capacity of a TDMA system can be
calculated in terms of numbers of channels per MHz as given by N = (1/B)/K. Where
"B" is the bandwidth per channel in MHz and "K" is the cluster.

face="Arial" size="3">Analysts predict that by the year 2000, over 10 percent of
all lines being installed will be wireless. Some other analyses say that by the year 2005,
number of wireless lines installed per year may overtake wired line.
SIZE="5">


And in the case of CDMA system, the
capacity can be calculated as N = (G/SIR)(1/a)(f.h.p). Where "a" is the voice
activity factor, "f" is the inter-cell interference, "h" is the
handover loss, and "p" is the factor relating to power control inefficiency. In
case of WILL systems, handover loss will be 1 as there is no mobility.

Typically, for a CDMA system, the capacity
is around two times the capacity of an equivalent TDMA system. However, by using Dynamic
Channel Allocation (DCA) TDMA, the capacity benefits of CDMA reduce to 1.4.

In DCA system, each cell has the access to
all the frequencies. Before transmitting, each cell measures the interference on each
frequency and dynamically allocates the channel with the lowest interference. This enables
TDMA system to use the same frequencies in all the cells.

It should, however, be noted that although
CDMA-based systems are having more capacity than TDMA systems, it is difficult to have a
concrete comparison, until there is widespread deployment of CDMA systems. Moreover, in
rural areas or low capacity areas, high-system capacity may not be of great concern. Thus,
considering all the factors including cost and the business potential, the choice can be
made between the technologies.

Network Roll-out Strategy

The network roll-out, after selection of
technology, starts with radio coverage planning, selection of the number of cells,
selection of cell sites, connecting cell sites with main network, deployment of microwave
links or leased links for interconnection, and interface with switching and backbone
systems.

alt="https://img-cdn.thepublive.com/filters:format(webp)/vnd/media/post_attachments/be868bcbbbe961de02029a33fd6b9ad403ebdfa8417d9c921296acdf86a2db7b.gif (19855 bytes)" align="right" hspace="6" vspace="6"> face="Times New Roman">Number of Cells: Selecting the number of cells is fundamentally
based on the business plan. The key inputs from the business model are the number of
houses in the area, density of houses, expected penetration, traffic per subscriber, and
the projected growth.

Depending upon the subscriber traffic and
subscriber density, the number of traffic channels can be calculated. The number of calls
required for capacity is given by Number of cells = (Traffic channels required/Traffic
channels per cell), where Number of channels = Number of subscribers x Busy hour traffic
per subscriber.

Number of channels is also linked with the
call blocking rate. For a particular blocking rate and erlang traffic, the number of
channels required can be determined from erlang B table. The erlang formula is :

Where P face="Times New Roman">B is the blocking
probability, A is the offered traffic in erlangs and N is the Number of traffic channels
available.

Selection of Cell Site: Site selection
can be based on two factors. Coverage area and capacity calculations. In case, if more
coverage is required in low subscriber density area, high buildings should be selected for
micro-cells, whereas in case if high capacity is the criterion, low height buildings
should be selected. Radio coverage can be planned by using proper planning tools. WILL
systems are highly based on Line of Sight (LoS) model. The general inputs for these
planning tools include:

Advertisment