1

Abstract--In this paper we revisited a methodology for planning

and capacity evaluation of community-deployed wireless mesh

networks in rural areas, based on radio equipment parameters for

WiMax technology. The capacity is described by the maximum

end-to-end transmission rate (throughput) provided to each node

that composes the network utilizing interference-based link

scheduling. A community-deployed scenario for Internet access is

studied utilizing 13 sites in rural communities located in the north

center highland region of Nicaragua in Central America. To

reduce the Internet service cost shared common access points are

used in mesh configuration under asymmetric traffic demands.

The radio propagation environment is estimated utilizing free-

available tools that can be used to support rapidly-deployable

broadband networks. Finally, the (upper-bound) capacity

resulting from this approach is computed by applying nonlinear

optimization, and the results show a substantial gain in

comparison with Wi-Fi networks operating on similar conditions.

Index Terms—Rural Communications, Wireless Mesh

Networks, Community-deployed networks, Wimax.

I. INTRODUCTION

ICARAGUA is the largest country located midway across

Central America, but one of the least densely populated.

Like its economic indicators, Nicaragua’s fixed-line

teledensity and mobile penetration is one of the lowest in the

subregion. Since 1990s important policy reforms has been

done by countries of Central America. The reforms have

fostered private sector investment in telecommunication

services and infrastructure providing benefits to their citizens

who now, at least in urban areas, have access to several

telecommunication services. However, although we have

observed tremendous growth in infrastructure and services

(voice and data), it seems to be economically difficult to

extend these benefits to rural areas without the financial

support of the government [1]. So far, the rural areas far from

the main cities have been unattractive to private investors due

to low population density, long distance, and usually irregular

terrain together with low incomes of potential users. In par-

ticular, this problem continues to be a challenge to the

governments of developing countries, whose ultimate goal is to

eliminate the digital divide.

This work was supported in part by the Faculty of Electrical and

Computer engineering (FEC) from UNI, Managua, Nicaragua.

In countries with so many needs, one may think that

investment in telecommunications might have low priority

level. However, for instance the Nicaraguan government

recognizes that the development of and access to

telecommunication means is a key enabler [2] in promoting

job creation, knowledge-based growth, business innovation,

access to valuable information, and can be utilized to improve

education and health-care assistance. Fig.1 shows the general

vision of the Regulator of Nicaragua (TELCOR). To guaranty

universal access to telecommunications to all citizens of the

country requires reducing what is called by the regulator the

market efficiency gap and the real access gap (aimed to be

financed by the Telecommunication funds called, FITEL by its

name in spanish). Furthermore, since the broadband market is

extremely very low in Nicaragua, with less than 1%

penetration. Recently, TELCOR has announced plans for

developing universal broadband access as well, in order to

boost the economic growth of the country. Broadband

technology for rural communications seems to be a key

element in this strategic planning.

Fig. 1. The general vision of the Nicaraguan Regulator regarding universal

access (version of the original in Spanish, www.telcor.gob.ni).

In this context wireless mesh networks (WMNs) are an

appealing technology to provide broadband rural area

communication. WMNs are those networks that mix the two

topologies of wireless networks, Ad-hoc topology and

infrastructure topology. The used of this technology could

influence important changes to the current mechanisms used

regarding the financial support done by governments. By

Wireless Mesh Network Based on Wimax: A

Broadband Technology for Rural

Comunications Oscar Somarriba Jarquín, Member, IEEE

National University of Engineering (UNI), Managua, Nicaragua

Tel:+ 505 22781460, e-mail: oscar.somarriba@uni.edu.ni

N

2

taking advantage of the self-organizing capability of mesh

networks the communities themselves or even new potential

local players can take an active part in the solution to their

own needs of telecommunications. However, even in the

foreseen scenario where potential users are assumed to be

fixed, urge the availability of easy-to-use and free-of-charge

tools for network planning and capacity estimation as the one

presented in this paper.

Recent work on this area have make used of simulation tools

to analyze the performance in Wimax networks like the one in

[4], and [5], however their work has focus on the performance

analysis and testbed rather than a common frame work to

facilitate community network planning. The contribution of

our work is to apply a methodology of planning for

community-deployed mesh networks making, this time based

on Wimax (Worldwide Interoperability for Microwave

Access) technology. By the way, Wimax is the commercial

name of the IEEE 802.16. Besides, a Wimax network has been

rolled out with an initial coverage over our capital, Managua,

in order to provide internet services during the fourth quarter

of 2009. This network works mainly based on the

infrastructure mode suitable for urban environment,

accordingly to the cellular paradigm.

On the other hand, WiMAX shares Wi-Fi's strength of being

a wireless technology and of allowing for even cheaper use of

unlicensed spectrum. Besides, WiMAX provides additional

range and better quality of service [6].

So, we will utilize of the free software called ”Radio

Mobile” that incorporate topographical information (with free

available cartography) combined with the use Google EarthTM

[3]. Using these tools we will present an approach that could

be used for the network deployment and then we will revisit

our methodology [1] for capacity evaluation. The capacity is

described by the maximum end-to-end transmission rate

(throughput) provided to each node that composes the

network. We denote the throughput by and N is the number

of nodes in WMNs, respectively.

The remainder of the article is organized as follows. In

Section II we present the system model and methodology we

apply to study the capacity analysis of a rural region in our

country. Next in Section III, we then describe the community-

deployed networks and capacity evaluation of them. Finally in

Section IV, we have the concluding remarks of the study.

II. SYSTEM MODEL AND METHOD OF ANALYSIS

We will present as an illustrative example the deployment

scenario for Internet access for 13 sites in rural communities of

Nicaragua introduced in [1]. These communities are located in

the north central highlands region of the country, at the

departments of Estelí, Madriz and Nueva Segovia. Fig. 2

shows the geographical locations for the communities of

interest utilizing a digital map of the terrain.

The general research question of interest that we would like

to answer is: What is the capacity that a community-deployed

MWNs can provide subject to low cost deployment constrain

for internet access for the 13 telecenters utilizing Wimax?

For low-cost user-deployment we consider the following:

• Share internet access: for low-cost internet service (network

service recurrent cost) nodes share access provider.

• Utilization of radio equipment parameters on the 2.4GHz

frequency bands (WiMax physical layers). Similar rules to the

FCC part 15 have been adopted in Nicaragua through the rules

in AA001-2006.

We follow a similar approach to the one presented in [1] but

also considering the use of mesh networking with parameters

for WiMax technologies similar the ones utilized in [7] but at

20MHz (bandwidth), in the unlicensed 2.4 GHz frequency

band.

A. Capacity Evaluation

To share common Internet access points in a mesh

configuration we assume asymmetric traffic demand from each

node to a gateway node connected to the internet and the other

way round.

In addition to that, we assume that the average traffic from a

node to the internet is 10% of the traffic from the internet to

that node. This is a reasonable assumption if for instance the

gateway is connected via an ADSL (Asymmetric Digital

Subscriber Line) service over the PSTN (Public Switched

Telephone Network) and we share this connection providing

similar capacity to all the rural telecenters.

The user-deployed scenario is analyzed estimating the radio

propagation environment of located nodes. The path-losses in

the network are derived utilizing the digital map GTOPO30

with the Longley-Rice model as implemented by the

simulation program Radio Mobile [3].

The (upper-bound) capacity resulting from a user-deployed

approach is evaluated by finding the link transmission

schedule applying nonlinear optimization [8].

Fig. 2. Rural communities of Nicaragua involved in the study (120 km

80km). Source: Map was derived using the freeware software “Radio

Mobile” [3].

3

B. Analysis and User-deployed Methodology

We can summarize the methodology applied in the next

sections by the following steps from [1]:

I) Network Deployment (User-deployment):

i. Determine the radio network parameters in use

ii. Define the Internet gateway to be used

iii. Determine the link budget between station including

the gateway (this is done with Radio Mobile)

iv. Establish the current network topology

v. If there are nodes that cannot reach the gateway, add

digi-repeater nodes (digipeaters) to extend the

connectivity by multihopping and go to step Iiii.

II) Capacity Analysis Deployment:

i. Define the external traffic load demand from each

node to the Internet gateway and conversely

ii. Calculate the routing matrix

iii. From step IIi and IIii estimate the link traffic load of

the network

iv. Compute the equivalent path-link routing matrix

Rmesh to assess the real relative traffic load

v. Determine the network capacity by interference-

based scheduling (STDMA).

At the early stage of the user-deployment, the network starts

composed only by the nodes located at the communities. Then,

progressively from the gateway towards the communities, the

users add digital-repeaters (if that is necessary) to connect the

network.

Next, to determine the network topology we assume that the

hardware utilized has the physical layer parameters shown in

Table I. Those parameters are based on utilizing equipment

with similar physical layer to IEEE 802.16-2004 group (IEEE

802.16d) operating on the unlicensed 2.4GHz frequency band.

Mesh Mode is an optional topology for user-to-user

communication in non-line of sight IEEE 802.16. It is included

in the standard to allow overlapping, ad hoc networks in the

unlicensed spectrum and extend the edges of the wireless

range at low cost. Mesh support has been extended into the

licensed bands too.

Moreover, to find the path losses, we assume utilization of a

20MHz channel on IEEE802.16d that may correspond to

operate on the frequency range: 2401 MHz- 2421 MHz. To

determine the radio propagation path losses we have used the

Radio Mobile freeware software with 20 meters antennas

height in the nodes, and higher antenna heights in the Internet

gateways that will be introduced in the following sections.

C. Traffic Model

We assume that the average from node to the internet is

10% the traffic from the internet to the nodes. This seems a

reasonable assumption if for instance the gateway is connected

via an ADSL (Asymmetric Digital Subscriber Line) service

and we would like to share this connection providing similar

service to all telecenters connected to the same gateway. In

ADSL over Plain Old Telephone Service (POTS), ITU

G.992.1 Annex A standard, the downstream rate is 12Mbps

and the upstream rate is 1.3 Mbps (which is 10.8% the

downstream rate).

TABLE I

WIMAX PHYSICAL LAYER PARAMETERS

D. STDMA Scheduling for Mesh Network

To find the interference-based scheduling for constant

transmission power and variable rate systems we follow a

procedure similar to the one described in [1]. Using the

available rate sets and required SINR in Table I we find the

sets of cliques containing links having the property that all

links in the same clique can transmit simultaneously selecting

one of the available data rates.

The columns of S can be linearly combined to create the

STDMA (Spatial TDMA) schedule by defining the vector of

weights = [1 …K]T corresponding to the fraction of the

time that each column vector of S is activated within a

STDMA frame. Hence, for a given the allocated link

capacity, c = [c1 … cL]T is given by

(1)

4

The capacity allocation can be done through a scheduling

algorithm. The scheduling algorithm allocates slots and

transmission rates depending on the amount of traffic passing

though each link and the objective function to be maximized.

In order to find the interference-based schedule for max-min

fair allocation we utilized the column generation method [8],

for constant transmission power and variable transmission rate

as formulated in [9] but with the redefined mesh equations

derived in the paper [1]. That is, we find the columns S, and

mesh [1…p] that solve the following optimization

problem:

maximize min

subject to i ≥ min i

Rmesh mesh S

k 1

min ≥ 0 mesh ≥ 0 0 1

III. USER-DEPLOYED AND CAPACITY EVALUATION

FOR RURAL COMMUNITIES IN NICARAGUA

A. Description of User-deployed approach

We assume a simple user-deployed behavior. The

community starts by locating their own nodes at their locations

(telecenters) and the shared Internet access node (gateway)

located at the closes land-based infrastructure. At the

communities under consideration there are two municipalities

headend (towns) where it is possible to have access to land-

based Internet services: San Juan de Limay and San Juan del

Rio Coco. The geographical coordinates are shown in the

Table II. In San Juan de Limay the operator Claro ENITEL

has a 63m height tower, and we assume that a similar tower

could be utilized (or is available) in San Juan del Rio Coco.

By utilizing these gateways we can group the communities

into two potential user-deployed subnetworks as shown in Fig.

3. We call subnetwork 1 to the set of communities that are

geographically closer to ENITEL in San Juan de Limay, and

subnetwork 2 to the set of communities closer to San Juan del

Rio Coco. Since in this paper the procedure for community-

deployed approach is similar to our previous work, we will not

elaborate more see [1]. TABLE II

GEOGRAPHICAL COORDINATES FOR INTERNET GATEWAYS

B. Capacity evaluation for subnetwork 1

We apply the described method before together with the

above user-deployed network, and also utilizing

omnidirectional with transmitter antenna gain of 16 dBi and

beam-steering system with antenna gain of 23 dBi (HPBW

stands for Half-power beamwidth). The capacity evaluations

(maximum end-to-end transmission rate o throughput) for

Wimax subnetwork 1 are summarized in Table III:

TABLE III

CAPACITY EVALUATION FOR SUBNETWORK 1 ()

Antenna

System

IM [dB] Antenna

Gain [dBi]

Uplink

(rate/node)

Downlink

(rate/node)

Total

rate

Omni 2 16 115kbps 1.2Mbps 8.8

Mbps

Beam-

steering

(HPBW

30)

2 23 233kbps 2.3Mbps 18.0

Mbps

Beam-

steering

(HPBW

30)

5 23 517kbps 5.2Mbps 40

Mbps

* Communities: San Luis, La Fraternidad, San Lorenzo, El Carizo, El Ángel,

El Mojón, San Juan de Limay (gateway). IM Interference Margin.

Fig. 3. Subnetwork 1 and subnetwork 2. Source: Map was derived using

Google Earth. Google INC.

We can note that utilizing an interference margin (IM) of

5dB and Beam-steering antennas produces higher end-to-end

data rate. This is because when utilizing IM=2dB with beam-

steering antennas, the network topology changes with respect

to the omnidirectional case.

In summary for Wimax technology, by the simple user-

deployed approach for the subnetwork 1, the (upper bound)

allocated end-to-end downlink transmission rate is up to

6.22Mbps per node (telecenter) and for the uplink is up to

622.4kbps per node in comparison to Wi-Fi networks that

provides 5.167 Mbps (downlink) and 516.7 kbps (uplink),

respectively and they have studied in [1].

C. Capacity evaluation for subnetwork 2

Again, for the gateway at San Juan del Rio Coco we assume

63m tower height and all nodes in the network are assumed to

5

be installed utilizing 20m antenna height.

When utilizing omnidirectional antennas with the selected

points there is not fully connectivity for all nodes that

composes the network. The utilization of beam-steering

antennas with an IM of 2dB make possible to connect all the

communities to the Internet gateway. The increment on the

interference margin is done in order to allowed more resistance

to interference and by this way we remove links with low

transmission rate (we avoid them to be used for routing traffic

and to be scheduled). We apply the described method before

together with the above user-deployed network and also

utilizing beam-steering antenna systems. The capacity

evaluations are summarized in Table IV.

TABLE IV

CAPACITY EVALUATION FOR SUBNETWORK 2 (**)

Antenna

System

IM [dB] Antenna

Gain [dBi]

Uplink

(rate/node)

Downlink

(rate/node)

Total

rate

Omni 2 16 - - - Beam-

steering

(HPBW3

0)

2 23 488.5kbps 4.9Mbps 32.4

Mbps

Beam-

steering

(HPBW3

0)

5 23 - - -

** Communities: El Ojoche, San Marcanda, El Varrillal, San Miguel, Cristo

Real, El Jobo, San Juan del Rio Coco (gateway).

Hence, for this user-deployed subnetwork using Beam-

steering antennae, the resulting (upper bound) end-to-end

downlink transmission rate is up to about 4.9Mbps per node

(community) and 488.5kbps per node for the uplink in

comparison to Wi-Fi, that provides 3.9 Mbps (downlink) and

388.7 kbps (uplink) respectively (already studied in [1]).

IV. CONCLUDING REMARKS

In this work we have revisited a methodology for capacity

evaluation of user-deployed multihoping wireless network

(WMNs) based on Wimax technology.

The evaluation results shown by an example of planning and

capacity evaluation in rural communities of Nicaragua

demonstrate that mesh networking is an appealing technology

to allow the communities themselves or even new potential

players to take an active part in the solution to their own needs.

This work also shows the need to develop easy-to-use free

software tools coupled with mesh network products to make

the community deployment of mesh networks ubiquitous.

In this work we have extended the methodology proposed in

[1] to include Wimax physical layer for community-deployed

MWNs in rural areas with low density users, which allow the

perform analysis and design of these kinds of networks.

MWNs are an important alternative for fixed wireless

broadband access to cover remotely located places in rough

terrain. However, we have to be aware that in 2.4 GHz

frequency bands for Wimax we can find, among others, further

interference from Wi-Fi systems already operational.

To study the systems we have considered three possibilities:

a) omnidirectional antennas with gain of 16 dBi, b) beam-

steering antenna systems with HPBW of 30 and antenna gain

of 23 dBi with IM=2 dB and c) beam steering with HPBW of

30 with antenna gain of 23 dBi with IM =5.

The results obtained in our subnetworks show a substantial

gain in capacity utilizing Wimax technology in comparison to

Wi-Fi networks (IEEE 802.11g), even though we are using

roughly similar bandwidths (20 MHz versus 22 MHz). Also,

WiMAX provides additional range than Wi-Fi.

Finally, even though WiMAX is economically feasible

option for rural communications, it is still potentially more

expensive than Wi-Fi networks.

REFERENCES

[1] M. Sánchez G., O. Somarriba J., and J. Zander, "User-deployed and

capacity evaluation of multihop wireless networks: A Case Study for

Nicaragua," In Proc. 2008 8th Scandinavian Workshop on Wireless Ad-

hoc and Sensor Networks (ADHOC08), Stockholm, Sweden.

[2] Fondo de Inversión de Telecomunicaciones (FITEL). [Online].

Available: http://www.telcor.gob.ni/Desplegar.asp?PAG_ID=15

[3] R. Coudés. (2012). Radio Mobile Freeware by VE2DBE. [Online].

Available: R. Coudés. (2012). Radio Mobile Freeware by VE2DBE.

[Online]. Available: http://www.cplus.org/rmw/english1.html

[4] C. Thomas. (2011). Simulation of WiMAX Networks and Allocation.

[Online]. Available: http://cse.wustl.edu/Research/Lists/ Technical%20

Reports/Attachments/954/NS3%20Simulation%20of%20WiMAX%20

Networks.pdf

[5] J. Ishmael, S. Bury, D. Pezaros, and N. Race, "Deploying Rural

Community Wireless Mesh Networks," IEEE Internet Computing, vol.

12 no 4 pp. 22-229. 2008.

[6] Wireless Options for Providing Internet Services to Rural America.

(2008). [Online]. Available: http://www.cse.wustl.edu/~jain/cse574-

08/ftp/rural/index.html

[7] M. Marques, J. Ambrosio, C. Reis, J. Riscado, D. Robalo, F. J. Velez,

and R. Costa, "Design and Planning of IEEE 802.16 Networks," In

Proc. 2007 of the 18th IEEE International Symposium on Personal,

Indoor, and Mobile Radio Communications (PIMRC07).

[8] M. Johansson and L. Xiao, "Cross-Layer optimization of Wireless

Networks Using Nonlinear Column Generation," IEEE Trans. Wireless

Commun., vol. 5, no. 2 pp. 435-444, Feb 2006.

[9] M. Sánchez G., J. Zander, and B. Hagerman, "On the performance of

power and rate control in STDMA multihop networks with advanced

antennas," in Proc. 2008 IEEE International Conference on

Networking, Cancun, Mexico.

BIOGRAPHYH

Oscar Somarriba Jarquín, received his title of

Electronic engineer from the National University of

Engineering (UNI), Managua, Nicaragua, in 1989,

and the Technical Licentiate of Engineering degree

in wireless digital systems from the Royal Institute

of Technology (KTH) in Stockholm, Sweden, in

1996. He also founded and headed the

Telecommunications department at UNI from

1989-1992. His employment experience included

the position as chief of staff and maintenance supervisor of the INTELSAT

earth’s station in Managua, RF Design & Optimization Manager of cellular

and PCS networks (Claro Nicaragua), consultant work and training courses

for several national and international agencies, among them: The National

Telecomm Company (ENITEL), The National Power Utility (ENEL), The

Nicaraguan Geophysical Institute (INETER), COCESNA, BELLSOUTH,

LIO/UN, UNDP/UN, and the World Bank.

His research focuses on complex problems in cross-layer optimization and

scheduling in wireless networks. He has published several papers on multihop

ad hoc networks for rural-area networks and emergency communications.

Currently, he is a senior Researcher with the Master's Programme in ICT at

UNI.