CIUEM MICTI:Project Chapter

FMFI Voice Messaging Project in Mozambique:

The challenges, innovations and experiences to develop sustainable communication means in the rural communities

Project Co-ordinator:	Eng. Constantino Sotomane (sotomane[@]micti.co.mz)
Senior Researcher:	Eng. Jamo Macanze (jamo[@]micti.co.mz)

Introduction

The self-sustainable operation of the rural public access points in Mozambique represents the biggest challenge and the key-factor for the success of such community centres. Many factors can help the achievement of self-sustainability, being the massive use of the community centre resources by the community, a measure that can produce faster results and at the same time increase the use of ICTs by the rural communities. Bringing more people to make use of ICTs can be achieved by creating new services that meet their needs, at an affordable price. However, such services cannot be implemented the existence of a reasonable Internet access infrastructure.

The majority of the Telecentres provide people with services such as e-mail, Internet, telephone, fax, photocopy, document typing and printing. In some of them, there are available other services such as computer training, graphical services, TV & Video, and community radio. The level of connectivity and access to communication means are poor and the costs are very high. Around the Telecentres, there are several governmental institutions, non-governmental organizations, bank representations, schools, healthcare centers, and as well some small - medium scale economic agents and all require access to information in their daily activities. People living in theses zones, are characterized by a low level of scholarship and a high rate of emigration (mainly in the southern region of the country) to the mining industry, in South Africa. Based on these facts there is a need to seek for innovative and low cost alternative means access to information.

The “First Mile First Inch (FMFI) – Voice Messaging Project in Mozambique” aimed at researching and deploying Wi-Fi links for Internet connection through sharing links between Telecenters (having a relatively better Internet access solution, e.g. leased lines) and some institutions in the neighborhood, and implementing VoIP-based Voice Message services.

The Wi-Fi component addresses the First Mile issues, and is also expected to contribute in human resources capacity building, by teaching high school students how to build low cost Wi-Fi antennas and involving them in the “mass production” of the antennas to be used in the project. The voice message component, which addresses the First Inch, consists mainly in implementing a Voice Mail boxes solution for Telecentres.

The project implementation was divided into two main phases, namely the wireless links setup and the voice message services. In the first phase the team established contacts with the Mozambican Ministry of Education partner, in order to identify the school(s)/students that would collaborate in the project. Equipment procurement process was also started, being the task that observed delays, due to the fact that some key components needed to build the antennas (e.g. N-type connectors, pigtail, some coaxial cable) are not available in the local stores. An approach on dealing with the problem was the request to our partner, the Mereka Institute at CSIR in South Africa to procure the equipment and then the team travel to collect the equipment.

Putting the telecentre as small-scale ISP using can-antennas for wireless links made successfully possible to link local state entities (in Manhiça) and NGOs (in Chóckwe). VoIP based solutions were tested within the Edurado Mondlane University Campus and between the telecentres and the University.

Challenges regarding the policy and legal framework that allow the use of both the telecentres as local ISPs and VoIP based application still be work out. The telecentres need to be licensed to operate as ISP and a license to use the 2.4GHz frequency (at the moment were used free license available only for research propose). VoIP is not licensed in Mozambique and on of the objective of this project was to influence the policy makers.

Research Question

Background to the problem

One of the biggest challenges regarding Telecentres operation is the problem of auto-sustainability. One of the solutions to this problem is getting more people from the communities using the available services at the Telecentre, which can contribute to reduce the price of the services to the end user. One way to achieve a higher use of the Telecentre resources is to introduce new services that meet the communities needs (such as contents in local languages), and as well, to improve the existing ones. Also it is crucial the design of strategies to make these communities become aware of the benefits of the use of ICT resources.

Proposed solution

The installation of an enhanced Internet access infrastructure (such as the installed leased lines), which is a basic condition for improving services provision, involves higher operation costs. In order to find means to contribute to cover the additional costs, new added value services have been identified, as an effort to attract more people to the Telecentre. Such services include the voice messaging (e.g., voice mail), relying on Voice Over Internet Protocol – VoIP technologies, and the use of Telecentre as a Small Internet Service Provider – ISP, by the use of low cost Wi-Fi technologies.

The potential of the voice messaging service resides on the fact that it can largely cover non literate people, that constitutes the majority in rural zones in the country, by allowing them to use their own local languages (without having to type a letter) to exchange messages with their loved-ones and relatives. Additionally, real time interaction is not required anymore, in such cases. The use of VoIP will contribute to lower the operation costs.

The use of Telecentre as a small ISP, will provide the surrounding institutions with Internet connectivity, sharing the same connection as the Telecentre, which means extra financial resources, and also a contribution to disseminate ICT s in rural zones.

Methodology

The proposed objectives of the project were to:

• implement innovative and low cost communication technology solutions that contribute to a major involvement of the communities in the use and development of ICT resources;

• facilitate and stimulate the development of local contents; and

• empower the local communities by the introduction of new ICT services and technologies.

In order to achieve the project objectives, the following methodology was used:

• Desktop research: Literature review and Internet search of VoIP and Wi-Fi technologies and equipment including guidelines for homebrew antennas, survey of regulatory aspects for the case of Mozambique. Also the reports of the assessments that have previously taken in Telecentres, regarding services and technologies in use and the required services by the user will be reviewed.

• Identification of pilot locations: This activity consisted of identifying the potential locations and establishment of agreements in order to perform the work at the chosen places. For locations were identified as the potential targets, namely:
  • the Manhiça and Chókwè Telecentres – public access points, in which will be implemented the home brew antennas to expand the Internet connectivity to some neighbourhood organizations,
  • the Industrial Institute of Maputo – a technical educational institution, in which there is a great possibility of getting students involved in the manufacturing of Wi-Fi low cost equipment (the homebrew antennas),
  • the Eduardo Mondlane University main campus – as the test bed for the preliminary tests before the implementation phase, in which the research team performs the tests, the research hub.

• Equipment specification: A detailed list of equipment to deploy the Wi-Fi and VoIP infrastructure that best sweets and meets the needs, both in terms of costs and technical performance, of the sites selected was elaborated. The acquisition process through the local and foreign dealers was follow.

• Implementation: The deployment of the proposed infrastructure at the identified locations took place starting by developing and testing the applications at the research hub. Also, the team delivered some train the Telecentre operators on how to manage the installed infrastructure and services.

• Tests and monitoring: Real tests were performed before starting providing the services in a commercial basis to the community. After that, a monitoring stage started in order to eventually assist the Telecentre operators, and as well see the impact of the installed infrastructure.

• Project results report: Periodical progress reports were submitted to the “First Mile, First Inch” project manager. Gradually, during the project, the documented activities and research information were compiled in order to produce the project final report including this project paper.

To guaranty the successfully implementation of the project and a share of lessons learned, the research team participated as a member of a regional research network that shared the challenges and lessons learned with other members. Sharing avenues were created such as workshops, teleconferences, e-mail discussion lists.

Implementation process

The diagram presented below, summarizes the “First Mile, First Inch” technologies implemented in the ambit of this project.

Fig.1. Implementation Scheme

Voice Mail – VoIP technologies

The team planned to install and test IP-Telephony to be used to make calls among the locations, voice messaging related services for the communities in each site, and as well Internet service provision from the Telecentre to the surrounding institutions, using low cost wireless connections. Also it was planned to test the Fax over IP service between CIUEM and Manhiça Telecentre. In order to accomplish this, a Digital Residential Gateway (DRG) device was used to interface normal telephone handsets and Fax machines, to the packet switched network infrastructure. Also, software based telephones were used to test the telephony service. The team installed a VoIP PSTN gateway, to allow calls between the private networks and the public telephony network, only as an experimental service as it is likely not to comply with the existing telecom regulatory aspects. VoIP Gatekeeper and other voice messaging related server software will be installed on server machines, acquired to provide store and forward and Internet Proxy services for the Telecentres. The priority goes for free open source software.

Fig.2(a)IP-telephony testing sessions with the research hub using the Digital Residential Gateway as interface and normal phone handsets and Asterix software

Fig.2(b)IP-telephony testing sessions with the research hub using the Digital Residential Gateway as interface and normal phone handsets and Asterix software

The test bed was the sub-phase, were the team studied and tested the technologies in a laboratory environment, before using them in a production environment. It consisted of the following activities:

• Technologies study;

• Installation and test of Asterisk in the Laboratory;

• Testing the Integration of the Telecentres in the VoIP system deployed at UEM main campus.

For the real-time voice communication tests, the team decided to proceed setting up an independent system and Asterisk was installed on the Telecenters server. In some technical aspects, asterisk seemed to perform well. After the laboratory tests (using Fedora Linux), Asterisk was successfully installed and is up and running on a real Server since January 2006. This server, which is the same machine that provides Electronic Mail service for the three Telecenters using leased lines, runs Unix FreeBSD 5.2 (this is emphasized due to the fact that FreeBSD is not considered Linux, the most common platform where Asterisk solutions are implemented).

Some tests such as Echo, VoIP Outbound calling (to users in other SIP server), voice menus, audio streaming and VoiceMail have been successfully performed between Maputo and Stockholm, where the one team member responsible for setting up the server was at the moment attending a three month training. Real-time communication tests are still experience some problems of cut-offs. Apparently this is not related to bandwidth constrains, since when one of the involved parts “holds on” the call, the “music on hold” function, in which an mp3 file is broadcasted from the Asterisk server. At the moment the server is being configured in order to reach and be reachable by other servers.

In case the tests reveal again that it not possible the voice communication over the 64kbps lines, the team has considered studying and implementing a different design, based on the concept of store and forward. Such solution will consist of independent asterisk servers running on the Telecentres proxy servers, instead of the centralized solution. It means that some mechanisms to allow one server to forward a voice massage in case it is not for a local destination should be identified.

WiFi Technologies

Fig.3.(a) Wi-Fi technologies for low cost communication in rural areas, a can antenna (directional) in use within the University Campus.

The second main test is the Internet Services Provision, through the use of Wi-Fi technologies. Wireless Access Points were installed in Manhiça and Chócwe Telecentres, so that clients using homebrew antennas can point to it and gain access to the Internet, by sharing the same connection as the Telecentre to its ISP.

Fig.3. (b) Wi-Fi technologies for low cost communication in rural areas, an enhanced can antenna that enables the increase of the gain up to 20dB and increases the distance reached without using amplifiers.

During the testing phase the team mounted a workshop at the university premises, in which the homebrew antennas were manufactured and tested. This workshop was planned to be replicated in the Industrial Institute of Maputo, with the collaboration of the Ministry of Education, in order to provide students with tools that will allow them to manufacture the above-mentioned antennas. Unfortunately this was not implemented duo to some internal constrains on the disposal of the IMM facilities including access to the local prosy server.

(a) can antenna installed inside the building during the verification of the signal range.

(b) can antenna installed in the hall getting the signal from the telecentre.

Fig.4. Installation of can antenna at the Manhiça District Administration Headquarter

The work regarding setup of the Wi-Fi links was delayed due to the lack of N-type connectors, pigtails, the recommended coaxial cables and some specific WLAN PCMCIA cards, which were not found available at the Maputo local stores. The team contacted CSIR and asked for advice on suppliers so that the missing parts were ordered and bought in South Africa.

Meanwhile, the team established some contacts abroad and did manage get free of charge, some wi-fi equipment (Wireless Cisco bridges, yagi and parabolic antennas) and books on wi-fi, VoIP and Linux. The offered equipment arrived in Maputo on December 2005, and will be used in some installations. While it’s true that the equipment is suitable to setup the links in some locations, the objective of teaching students how to build and low cost equipment cannot be achieved.

(a) Replacing the can antenna at the Machiça Municipality (b) installing the wireless adaptor card in a computer at the HOPE and NGO in Chóckwe

Fig. 5. Installing wi-fi links fir internet access in rural communities

A series of test were realized within the University Campus to determine the performance of the homebrew antennas. For that it was used a tool called “netstumbler” and two different type of antennas to compare the performance being one acting as the reference antenna with a known gain. During the process, two different measurements are needed of both antennas (known and unknown gain) by using the monitoring software the highest values are the ones being considered and serve then as reference value.

(a) training session at the Manhiça Telecentres on configuring wireless antennas (b) interview with the telecentre manager in Chockwe to identify the best solution and potential clients of the wireless links to be implemented

Fig. 6. Site assessment and training sessions at the telecentres

Two organizations in Manhiça District (the District Administration Headquarter and the Mayor Office) and two NGO in the Chockwe District were connected to internet through the tecelecentre leased lines using the wireless links by the can antennas. To maximize the use of internet, MICTI provided for free a short training in how to use internet, temporary free access to internet. The distance between the sites and the telecentre is about 100-200m with a line of view allowing wireless links with about 80% of quality of the signal.

Challenges

During the implementation of this project the team were faced with various challenges:

• Human capacity: for the implementation of this project was used students which had to be trained to acquire the needed knowledge to implement and monitor the project. This process resulted in a somehow endless cycle due to the fact that the trained people often get better job opportunities with the acquired skills and end up leaving the project. The approach used was to involve as many as possible students so that when one left the group the others would continue the implementation.

• Procurement of equipment: Equipment procurement process observed delays, due to the fact that some key components needed to build the antennas (e.g. N-type connectors, pigtail, some coaxial cable) are not available in the local stores. An approach on dealing with the problem was the request to our partner, the Mereka Institute at CSIR in South Africa to procure the equipment and then the team travel to collect the equipment.

• Connectivity at the Telecentres: the existing bandwidth in the telecentres was not sufficient to provide a communication using VoIP based solutions, there were many cuts during the communication. Because of the only one component of the project (the wi-fi links) was fully implemented and it up to run at the moment.

• Project management: being a project involving many stakeholders somehow there was lack of coordination and building of the same vision of enhancing the ICT activities & services using low cost communication means.

• Policy and legal framework: the proposed solutions applied in the project as policy implications. The VoIP telephony is not allowed in the country, the fixed telephony still monopoly of TDM and can only be used in the areas were TDM is not operating. The team has only implemented this component for research proposes and is not allowed to deploy this technology. The small scale ISP being implemented in the telecentres faces to constrains being the one related to the use of radio frequency that needs to be authorised by the communications regulatory authority and the second related being a service provider that needs also a licence from the communications regulatory authority. This is an impediment and increases the costs of operations and affects the sustainability of the project.

Lessons learnt

The major lessons were identified within the major challenges faced in the project.

• One of the objectives of the project was to massificate the use ICTs, a partnership with the faculty engineering of the Eduardo Mondlane University made possible to continually source students to work on the project for at least 6 months were they gained practical knowledge on networking and VoIP telephony while helped MICTI to implement the FMFI project. Portuguese literature were developed being one the guideline on how to build can antennas and other a linux systems administration course manual. These material were used to train the new members that were joining the project.

• The can antenna is a new concept in Mozambique, a lot of advocacy is needed to take place to demystify about the use of technology for development. MICTI has participated in various events as speaker or exhibitor presenting about the potential of can antennas. This also helped to demonstrate the MICTI capacity on implementing ICT for development projects and build a close relationship with the regulator authority in Mozambique.

(a) eng. Z. Nhacumba explaining about the can antenna to the Minister of Scienece and Technology (left) and the CEO of the Critical Software during the exhibition at the launching of the programme “Moçambique Tecnológico” in Maputo. (b) From left to right eng. A. Saica, Ms D. Salomão, Ms. J. Pondeca and eng. Z. Nhacumba at the MICTI stand during the 4th Science and Technology Fair held in Maputo

(c) eng. C. Sotomane explaining about the research activities at MICTI to H.E. the President of the Republic of Mozambique, Mr. Armando Guebuza and the benefits of the can antennas implemented in Manhiça and Chóckwe Telecentres during the International Conference on S&T for Development held in Maputo.

Fig. 7. MICTI advocacy on technology about the potential of can antennas

• It was crucial the close trust relationship with Mereka Institute which helped MICTI in equipment procurement in South Africa. Using this approach, MICTI is now discussing with Mereka Institute to sign an MoU were will include other research areas such as Open Source, mobile applications and mesh networks.

Implications for practice and policy

The implementation of this project took place at the district level, appointed at the Government’s 5-Year Plan as the focal point for development. Appling this solution helped the district government administration and the municipally having access at a reasonable price at the same time contributing to the sustainability of the community centre which would then continue providing internet access to the public in the district. This approach used by the project members brought a challenge to the policy makers in terms of thinking on which framework would be used to regulate the use of communication means in disadvantage areas were the business market still infant.

MICTI was invited to participate on the discussion forums for the development of the Universal Access Policy and its implementation strategy. Lessons learned from this project were shared with other experts in this forum.

It is only possible to implement this project at legal level being a research project and not as a commercial project as a licensing is only free for research propose. Much work has to be done to allow the introduction at a commercial level to allow the creation of local entrepreneurs providing such type of services in rural communities at a reasonable price.

Monitoring and evaluation

Before the implementation of the project an assessment was conducted at the telecentre to identify the user needs and means to build a business model that would enable a sustainable operation of the telecentres. This assessment identified two key areas being first a solution that could reduce the cost of telephone calls and the second on extending the internet access to more people in the community. A research on innovations that could respond to this problem was undertaken and applications developed.

An Outcome Mapping Methodology was implemented monitor the implementation of the project focused on measure the changes in behaviour of the direct stakeholders/partners (boundary partners) involved in the project. An intentional design of the Outcome Mapping was developed to collect information through periodic outcome journals using interviews, trip reports and other information made available by the partners. A set of periodical visits took place to the boundary partners.

The identified boundary partners, shown below for the Voice Messaging Project are presented in levels of interaction and influence. For starting point the project intends to target individuals and then target institutions:

• The Industrial Institute of Maputo (IIM) and the faculty of engineering of Eduardo Mondlane University is the focal point for knowledge and technology transfer. Target entities are teachers and students.

• The Telecentre is where the project is seeking to build business model and use as a dissemination unit of the FMFI technology in rural areas. Target entities are managers and users.

• The Mozambique National Institute of Communications (INCM) is where the project itends to influence the telecom policy legislation process to accommodate FMFI technology for a rapid rural connectivity in a low cost basis.

• CSIR - Meraka Institute and IDRC are the major supporting organizations for the development of this project.

Below are presented some changes on behavior to the key parties, the people who benefited and the regulator responsible to create regulatory environment, involved in the project implementation that were verified during the implementation of the project:

The Telecentre

• The Telecentre manager attended training sessions of how to install home brew antennas and on how to make first line maintenance and making first line maintenance to the clients linked to the telecentre. They are now able to assist their clients who are having access to internet and provide training on how to use internet.

• Telecentre managers meeting with diferrent local organizations to source partners that would contribute with resources to sustain the telecentres. Services agreements were signed with the some surrounding organizations.

• Research in the internet being conducted by the telecentre managers/staff in order to provide the community with information on health, education being published at the community radio attached to the telecentres.

• The Telecentre users benefiting with the wireless links able to use internet on their work. For example in Manhiça there is a commission responsible to write the district strategic development plan used internet to research on how to write strategic plans and got remote assistance from experts living in Brasil and Portugal on writing and revise the document. Another example is the Municipality which used the e-mail for to do the procurement (requesting and reception of quotations), then reducing costs on the sending faxes or travel to collect the quotations. They are considering on developing a web site to publish activities undertaken and business opportunities within the municipality.

• The District Government Administration is open to provide personnel to be trained to work at the telecentres and the technical officers. This would contribute to retain the staff at the district after trained as they would have a permanent employee contract.

The Regulator

• The regulator has received the research team and presented the worries related to regulatory aspects and has visited the project in Manhiça.

• Regulator invited MICTI to attend some working meeting for the development of the Universal Access Policy and its implementation strategy and the MICTI project was identified as innovative idea that could be used by local entrepreneurs to provide access in remote areas.

• It was established an interaction channel between the regulator and MICTI to develop and implement innovative communication solutions for communities, frequent meeting held between parties

The faculty of engineering

• There is a competition among students to come to MICTI to make their internship as an opportunity to learn and implement practical projects, put hands on technology and had opportunities to meet and discuss with several experts.

• Most of the students who participated in this project gained knowledge who facilitated them to get better jobs, two are now working to a software company in Portugal, the other two are working to mobile service provider.

• The research skills gained were useful for two of them who went for master’s studies on software development and networking.

A monitoring scheme was also setup to monitor and evaluate the progress of the project based on the accomplished of the proposed activities. The team believes that some management decision definitely contributed to the positives results of the work. One point to be outlined is the established partnership with CIUEM and the Faculty of Engineering, which produced good results given its technical expertise and wide experience in IT field. Additionally, the previous knowledge the team acquired during the design phase, was one of the major strengths to the project, since it provided a wider overview of the available technologies and reduced the dependency in terms of technical support relatively to the services providers and other entities.

The decision to go for a local applications development model in which the team had to build own antennas and develop VoIp telephony applications instead of relying on market equipment suppliers implied a higher initial investment. However, it gave the team a full learning opportunity to for build antennas and get knowledge on radio frequency technologies and linux systems administration, which would be difficult in case the antennas were owned and administered by the supplier.

Research Findings

The findings of this project can be seen in two components, at the technical level and at social level:

Technical Component

Homebrew antennas guidelines were developed n Portuguese as a result of the learning process that the team went through to implement the wi-fi links in the telecentres with the surrounded organizations. A solution was implemented in the telecentre that contributes to the financial sustainability of the telecentres as the organizations pay a fee for the internet link.

A documentation process on the VoIP telephony was conducted and can be used in future to continue the research to a solution that can be implemented with the existing bandwidth limitations.

To link the sites it was needed to use wireless routers, as they are expensive and not available in the country were used refurbished computers for the same functionality.

The students involved in project published two thesis on building low cost communication technologies. This research reports provides guidelines and test reports that would contribute to build new antennas with longer ration of coverage. Two new types of antennas were developed, one Omni-directional and other a directional resulting from a combination of can and satellite TV dish.

(a) Omni-directional antenna installed at the university campus for testing and determinations of the gain. (b) Combination of two technologies the antenna using the can to collect the electromagnetic waves and the parabolic reflector to increase the gain Fig. 8. Improved antennas

Social Component

The team has participated in some promotional events, one exhibition organized by the Ministry of Science and Technology and a dedicated TV section at the National TV (TVM) to speak about the homebrew antennas. This was used to demystify about the use of technology. With the guidelines people are able to build their own antennas.

The students who participated in the project had contacted with and access to various experts who contributed to they skills development and now some had access to better jobs in prominent ICT companies and have initiated their own business on networking applications.

A close relation was build among the telecentre staff and the community radio. The local community is benefiting with the news researched in the internet including content on agriculture practices and advocacy on HIV/AIDS and malaria.

The district government benefiting with the access to internet in their facilities are willing to provide a staff to be trained and then assign then to the telecentre to provide technical assistance, then locally the district have access in short time maintenance when the internet link is down.

Building a sustainable model

An assessment made to the telecentres identified at the bottleneck in their operations is the cost and quality of the connectivity. A solution was to bring a concept that could allow the enhancing of the quality of the link and the same time putting in place a mechanism to cover the costs of that link.

The solution then proposes that the telecentres would implement wi-fi links to the surrounding organizations with capacity to pay for an internet link. The monthly fees collected from these organizations would be used to pay the telecentre ISP. As more organizations request to be linked more bandwidth the telecentre would request to the ISP as this will be on shared basis.

With the aim of building local entrepreneurs who then could technical assist after implemented the solution in the districts, a concept of homebrew antennas was used which includes lot of technology and skills transfer to the people involved in the project implementation and there is an opportunity for learn by doing and to run open courses on homebrew antennas. The replication of this initiative is present has during the implementation new people are trained and involves local organizations (the benefits), local entrepreneurs (the future service provides).

The rural communities benefited by having access to internet in a reasonable price from more places, local government workers are having access to internet for free in their jobs.

Conclusion and the way forward

Conclusions

• As the overall assessment of progress made it can be said that this project is satisfying the defined objectives, despite some constrains happened with the equipment acquisition and frequent changes in the research team. These constrains delayed the implementation of the project, but the efforts and engagement of the new team have helped to compensate this gap created and now we have built the antennas and made developed a new concept for the VoIP services to be implemented in the telecentres.

• The low bandwidth allocated to the telecentre is one of the barriers to implement some of the services proposed. In this regard some modifications had to be done in regard to the VoIP based services and reduced the number of sites to be linked to the telecentre.

Recommendations and the way forward

As the main goal of the project is to propose a business model to sustain the telecentre operations, the revision on the project has to take into account issues related to:

• Number of sites connected to internet through the telecentres and the bandwidth allocated to the telecentre and to each site

• VoIP based services that not require high bandwidth.

• Most of the people living in the in the neighborhood of the the telecentre are not technology literate, then it recommend the implementation of user-friendly technologies