Login | Home | Story Behind | Press Info | Partners + Sponsors | Contact/Imprint 



 Friday, 03. September | Saturday, 04. September | Sunday, 05. September | Workcamp 6.-10. September 
last edited 7 years ago by admin
Print this page  

BMSTDA - Broadband Multi-service Switching, Transmission and Distribution Architecture

Genesis of BMSTDA requirements suitable for all alternative telecommunications networks

This is an excerpt from a report I wrote in March 2004, entitled "Draft Report on Rural Connectivity Planning And Related Locally Sustainable Technologies" as a Short Term International Consultant working under the ICT – Assisted Development Project/ World Bank P078458 for the Ministry of Capacity Building, Federal Democratic Republic of Ethiopia. If you have any queries about the content or requests for clarification, I can be contacted at the following address. or through Arun Mehta of Radiophony if needed who is attending the programs in Djursland.

Samudra E. Haque, S21X
Managing Director
PraDeshta Limited
House 24/E. Road 13/C
Dhaka, Bangladesh
Phone: +880 2 9882751, +880 2 9881126
Fax: +1 202 478 1998

(c) Samudra E. Haque. September 3, 2004. All Rights Reserved.
Author contact: haque at at nmc.pradeshta.net

In April 1999, several brainstorming sessions were held at the “Workshop on Internet: South Asian Realities and Opportunities” , in Dhaka, Bangladesh during which discussants from South-Asian countries created a set of recommendations which included two significant recommendations:

  • Newly emerging communication technologies and their convergence should be explored for enabling rapid growth of Internet service and infrastructure.
  • Explore and create new markets for Internet services, particularly in the rural areas (which are conventionally supposed to be non-profitable sectors), to make the establishment of telecom facilities viable and profitable in these areas.

In 1999, Wireless LAN technology (later to be known as IEEE802.11x standards) was in its infancy and was showing promise, but the prevailing method of establishing telecommunications services to any areas in countries such as Pakistan, India, Nepal and Bangladesh was to follow the “traditional” telecommunications architecture of centralized, switching, transmission and distribution facilities. This “top-down” methodology of network made it possible to provide services in quantity for those areas that had dense population demographics. It did not make it easy to spread communication services to sparsely populated, rural areas where heavily centralized services could not be rolled out economically. The common “standard” transmission technologies in use at that time were fixed microwave, fixed satellites, fixed optical over cable, fixed free space optics. The common “standard” switching techniques in 1999 did not consider IP as the primary method of either switching of transmission, yet it was obvious that the demand for Internet services was expected to increase rapidly at over 30% per year growth.

At the time the workshop was held, Pakistan had 40 ISPs in operation serving 250,000 users; Nepal had 4 ISPs with 8,000 users; Bangladesh had 18 ISPs and 40,000 users with India having a staggering 1,000,000 users with a dozen ISPs. Each of these countries was connected to the Internet in 1993, first by electronic mail through UUCP, then gradually by on-line satellite or fiber connectivity. The 1999 usage statistics were calculated by experts in the respective ICT industry who monitored growth and there was a clear consensus that this number represented only a small fraction of those who had access to (a) Computer Technology (ICTs?) and (b) public switched telephone network (PSTN). Wherever PSTN and ICTs? were available people were “connected”, otherwise there was no chance they would be until a Point-of-Presence of the PSTN would be established nearby.

The challenge was therefore to consider a technical solution that could possibly satisfy the needs of the “lower-income” population which would be typically found in rural areas and consequently deemed “unprofitable” by the established telecommunication companies. In addition, a low-cost, low-technology solution was not desired, through the brainstorming sessions it was made clear that there is always a technical solution to make products work more efficiently and at lower-cost providing innovation is rewarded for clever design and workarounds. Another issue that came up frequently was the requirement that the equipment be serviceable in the field, and that it be able to be repaired with the minimum of spare parts – a very demanding and difficult issue. Most of the practitioners in the field of ICTs? were frustrated at deploying communication networks and having to cater for the frequent breakdowns and repairs having had to been sent back to the factory. Network service quality and time to repair were very important attributes of any new technical solution as opined by the workshop participants. In fact the additional challenge was given to develop a class of technology that could be fixed in the field where there would be usually no developed infrastructure, a place that we would call “zero-infrastructure” area.

Throughout 1999 this idea was discussed after the workshop in several other venues by the various participants and this consultant decided to formulate a set of specifications for a new telecommunications network architecture that would fit all the requirements of the 1999 ICIMOD workshop. This specification was titled “Broadband Multi-Service Switching, Transmission and Distribution Architecture” which was abbreviated as BMSTDA and in its simplest form could be stated as:

  • BMSTDA communications technology solution must be quick to implement
  • BMSTDA communications technology must cost less than existing switching, transmission and distribution solutions
  • BMSTDA communications technology must be broadband in performance
  • BMSTDA must be independent of any particular transmission method in use
  • BMSTDA must be able to use hybrid communications technologies
  • BMSTDA must use a distributed, fault-tolerant, partial-mesh network topology
  • BMSTDA communications technology must make use of common-of-the-shelf parts and components as much as possible and should be interchangeable.
  • BMSTDA technology must be able to repaired in the field
  • BMSTDA technology should be built and designed taking note of the needs suitable for “zero-infrastructure” regions where it is likely that BMSTDA will be the first such communications technology to be introduced.

Prior to 1999 I had been involved in Research and Development of various types of data communications services for use in Bangladesh through the facilities of PraDeshta Limited. Amongst these projects one innovative requirement of a potential NGO was to use the video/audio output capability of a standard computer to playback recorded digital content for distribution of educational/entertainment material in the villages, through reception of satellite signals. The computer would have had a satellite receiver card installed to receive multicast data. This idea actually made technical sense, but the idea had to be modified to allow broadcasting over open airwaves on VHF channels in order to reach a wide audience of several villages who would tune their television sets to a miniature broadcasting terminal connected to the PC. The total cost of this setup was actually less than US$100 for an effective range of about 4 miles in diameter over open ground and the tests were demonstrated in PAL-B colour TV transmission broadcast standard. An important feature that had to be figured out was the minimum RF signal strength that would be required to broadcast a signal, yet be within the minimum legal parameters that are commonly allowed for any transmitter in use in Bangladesh. This “minimum legal power limit” is an item that is regulated on an individual basis by each country telecommunications industry regulator. It refers to the smallest signal power that the regulator will permit before considering it to be a broadcast and therefore would need a license. For example, the signal strength for cordless telephones is rarely above 1/20th of a Watt (or 50mW) while that of a cellular phone barely ½ of a Watt signal output at its maximum. By clever use of antennas and shortening the distance from the transmitter to the antenna even low power signals can be sent farther away than expected. Since BMSTDA architecture was being formulated at this time, the two technology tracks were combined and therefore two additional attributes of BMSTDA was defined thus:

  • BMSTDA networks must support multiple services
  • In case of radio links used in a BMSTDA network, the least amount of signal must be used in order to be legal as per local regulatory guidelines.
A model network to provide “proof-of-concept” of BMSTDA was created by this Consultant for commercial use by PraDeshta Limited in 1999 and the results of that experimental network was reported as a work-in-progress paper presented by invitation at COMMSPHERE2000, " Doing More with Less, or Deploying Broadband Communication Networks In Large Areas with Small Budgets" , http://www.tenet.res.in/commsphere/s7.2.pdf, Indian Institute of Technology, Chennai.

Some international news reports reviewed the technology and provided their comments in various articles such as:

The longest distance that was achieved using those early terminals was 16 kilometers line-of-sight at speeds of 3.5 Mbps. This network was replaced with a commercial service network connecting three points in Dhaka city which was operated until 2001, with another two distinct network of two and three nodes respectively operating from 2000-2002. There were at least four generations of devices (detailed pictures have been deleted from this WIKI, available on request from the author directly) that were experimented with:

  • Model 1: Initial 1999 developmental version, simple bare bones computer in conventional cabinet with expensive microwave low-loss feeder cable. For indoor/short-range use only; not for commercial service; Failed in outdoor tests. Antenna was simple indoor type. Commonly called “Broadband Wireless Router”.

  • Model 2: 1999-2000 “Yellow Box”, custom outdoor cabinet for mounting at Base of tower with expensive microwave low-loss feeder cable. For medium-range distance commercial service. Failed due to heat/cold problems. Range: 16+ Km @3.5Mbps duplex data rate with RF power output of 100mW. Electronics payload used conventional PC motherboard. Antenna was 24 dBi grid parabolic. Commonly called “Mast-Mounted Microwave Router Unit”.

  • Model 3: 2000-2002 “White Box”, custom outdoor cabinet with double insulated wall, fireproof, very heavy (25 Kg+ weight). It proved to be unwieldy for installation on top of tower. Range: 38+ Km @3.5Mbps duplex with RF power output of 100mW. The design was rejected due to extremely heavy weight problems at top of tower and lack of adequate cooling. Electronics payload was conventional PC motherboard. Needed no expensive microwave feeder cable. Antenna was 24 dBi grid parabolic. Commonly called “Mast-Mounted Microwave Router Unit”.

  • Model 4: 2002-2003 “Grey Box”, custom small outdoor cabinet with anodized aluminum heatsink and internal convection fan as well as sophisticated industrial strength Single-Board Computer and Passive Backplane. Reliability very good, distances achieved 22 Km at 5.5Mbps @100mW RF power output and 4 Km very good quality links at 30 mW RF power output.. For maintenance purposes, keyboard and monitor was attached directly to the router which was placed at the top of the tower, and the system did not need expensive microwave feeder cable. Antenna was 24 dBi grid parabolic. Commonly called “Mast-Mounted Microwave Router Unit”.

  • Model 4B: 2003-present. Similar to Model 4, but all connections to Keyboard and Monitors input/output were installed on the OUTSIDE of the cabinet, not the INSIDE of the cabinet thereby making it possible to seal the unit after construction, assembly, integration and testing and also to prevent water seepage. An additional benefit was that on certain units, we used low-cost DC power supplies instead of expensive Switched Mode Power Supplies to reduce weight and manufacturing cost further. “Commonly called “Mast-Mounted Microwave Router Unit” abbreviated as “MMRU”.

Along with the Mast-Mounted Microwave Router Units that were produced in an effort to validate the BMSTDA approach, a set of Antenna Tower Masts or Communication Towers were also designed and fabricated which allowed the build up of cheap telecommunications infrastructure including Antenna mounting kits for accurate AZ-EL pointing of microwave antennas.

The latest results from long term applications in the field are:

  • Distance: 22km across all weather conditions
  • Data Rate (IP): 5.5Mbps duplex
  • Traffic Type: Ethernet LAN-to-LAN
  • Max Number of simultaneous links: 4 (four); depending on hardware
  • Status of Weatherproofing: Adequate but not excellent

Next steps (in research)

  • Fully compliant weatherproof cabinet (low-cost): In service by October 2004
  • Integration of highly sophisticated small form-factor single board computer: Already done
  • Integration of high power microwave amplifiers for extended range and data rate: November 2004
  • Trial in harsh conditions: October 2004
  • Technology Available for licensing: Available free of charge to NGO or Non-profit organisations; Commercial offers welcome at nominal cost.
  • Training and Development Materials: Available at cost
  • Demonstration Kits: Available at reduced cost

Common Off-the-Shelf (COTS) peripherals for BMSTDA use

Choosing hardware that fits the requirements of BMSTDA network technology turned out to be easier than expected despite the ambitious list of objectives.

In the marketplace, dedicated Wireless Access Point hardware and commercial broadband microwave distribution networks were not considered for adoption, as they proved to be (at that time) simply too expensive, and they could not be manufactured or repaired in developing countries. In their place, software based routers such as Linux operating system based hosts with multiple network interface cards were considered and since the hardware foundation would remain a standard PC in all respects, almost any application could be adapted to run on it, in the manner that suited the particular application.

Therefore the following basic components were chosen for inclusion in BMSTDA component list:

Motherboard: Almost any computer motherboard will work, whether based on Intel’s chipset, AMD’s chipset or any other manufacturer of a motherboard as long as it supports PCI network interface cards, ISA cards (if desired), video/audio interface cards and is compatible with any version of a open-source operating system such as Linux. Integrated Motherboards are not preferred despite their absolutely rock-bottom prices since their failure rate is much higher in grueling environment. Therefore a choice was made to use Industrial-rated Single Board Computers and Passive Backplanes since that allowed for two major advantages:

  • Ability to withstand huge variations in temperature
  • Modular design and not integrated allowed for faster repairs/replacement.

Power Supply: Almost any power supply will work, as most Motherboards or Single Board Computers require only +/- 12Volts and +/-5 Volts and therefore these power supplies can be made by any competent EE student or graduate on a commercial scale. However there is a growing trend for adapting devices for use of Power-over-Ethernet connections (where the device obtains power from the Ethernet cable) so their may be a potential for further optimizing the power requirements by investigating just how little a device can be provided power for it to work. In the future, it is expected that custom hardware can be designed and manufactured that will have much simpler power requirements.

Software: Each BMSTDA device will have to perform functions in at least one of the following categories:

  • Dynamic Routing between local and wide area network interfaces,
  • Switching of data between its local area network interfaces,
  • Fault Tolerance Management, Operation and Maintenance functions,
  • Alarm Management and Reporting,
  • Event Management and Logging,
  • Multi-media Services,
  • IP Telephony Services,
  • Video Broadcast Services,
  • Audio Broadcast Services,
  • Network Discovery and Negotiation Services,
  • Proxy Services,
  • Authentication Services,
  • Firewall Services

So far to date, a recommended software application that provides some, but not all, of the desired functions above has been used successfully in the field, authored by Mikrotik of Latvia . There are now renewed efforts to build custom-Linux operating system that would provide a rich feature set but these are not at a point where they are significant or can be recommended. Some of these development efforts are noteworthy but ill-fated (such as http://www.linuxrouter.org/) and which have failed in the USA due to lack of interest by commercial companies who did not want a cheap, mass-produced router. In fact the software capability already exists with source code within that project, but ultimately it was abandoned because it was created in a market that was already developed and did not want to investigate low-cost solutions at that time. Perhaps a further study would be useful to see what other good efforts are available out there. Such an effort to ”acquire” technology would pay off in a short time as the “time to market” would be reduced by much less. See also http://doggerdog.bravepages.com/freesco/, where similar collaborative initiatives are reported to build routers and devices. A very popular implementation of a routing module for Linux operating system is the http://www.zebra.org repository of GNU routing software provided free of charge.

We assume that in the future the capability of computing hardware will be such that many other services can be performed simultaneously at the same time and this will need Research and Development support, but for now the objective is to make use of our limited skills and hardware capabilities at present.

Network Interface: The following types of network interface cards have been tested and are known to work with a variety of Operating Systems (including, non-Linux Operating Systems): Ethernet devices, IEEE 802.x spread spectrum devices, Fiber Optic devices, Free-space Optic devices, HF/VHF Radio devices, Dialup Modems, xDSL devices, Telephony Servers etc. As long as there is a compatible driver to interface the hardware with the software operating system, a BMSTDA device can make use of the data transmission capability of the device and “connect” to any other node with a similar networking arrangement. This proves the versatility of the BMSTDA in that it doesn’t care what device is connected to it, as long as it is able to communicate through that device and run networking protocols to manage traffic.

Memory: In order to run the Operating System, the BMSTDA device needs Read-Write Memory, and it must match the particular type of hardware used. With a custom-developed Operating System it should be able to be installed and made to work in a device with a minimum of 32 MB of RAM for basic routing functions but if more advanced services are required then 64MB of RAM upto 256MB of RAM may be required for the current generation of service applications.

Processor: Any kind of processor that matches the Single Board Computer, mainboard or motherboard and supports an open-source operating system can be used. The speed of a Processor will dictate the total aggregate data communication traffic that can be passed through each of its interfaces, and the higher the speed the better obviously. However careful measurement and planning of the (a) heat generated by the processor and its cooling requirements (b) the amount of processing power needed at present and in reserve for the future (c) calculation on the amount of dynamic memory consumed (d) crash recovery process if a processor exception causes total hardware failure. There are both professional and amateur ways to make sure a device stays operational even when hardware inside of its cabinet has failed; this is an important issue as a BMSTDA network is likely to be used for broad-based service to far-flung communities.

Flash Media: The storage device for the Router Operating System. Early on in the development phase of the MMRU hard drive technology was found to be notoriously unreliable for continuous service in varying weather and environmental conditions. Also, proper operation of a moving, mechanical device was subject to stringent safety and environmental guidelines that could not be guaranteed in the field. Therefore a decision was made to invest in the (more expensive but reliable) solid-state flash memory modules which interacted to the Processor and Operating System as standard hard-drive, but was actually a non-volatile device, and would retain its information for long time, without power.

Hard Drive: For Multi-media applications such as storage of digital content, application service, broadcast from stored files, Flash Media will be an inefficient storage device and will be ultimately slow in performance if repeated read/write attempts are done to it. In addition, the data storage requirements of multi-media will be often in the Gigabyte sizes which will result in cost-prohibitive Flash Media modules. Therefore for some of the BMSTDA configurations (see later section) normal Hard Drive installation is recommended, but only if the unit is kept indoors and in a controlled environment.

Internet Support Equipment: The local distribution of the network is achieved by a combination of any type of interface possible, and typically may include Ethernet Hubs, Ethernet Switches, Plain copper wire, fiber optic cable, Wireless Access Point etc. and the network can be expanded as much as required by duplicating the technology for each layer.

Mix-and-Match to make BMSTDA network devices

As stated earlier, devices of BMSTDA had to use common components as much as possible. Some configurations of the same type/class of function, hardware and software were frequently used and these were called:

  • Broadband Router (BR)
  • Broadband Wireless Router (BWR)
  • Mast-Mounted Wireless Router (MMRU)
  • Multi-Media Terminal Server (MMTS)
  • Micro-Community Broadcast Station (MCBS)
  • IP Telephony Gateway (IPTG)
  • Micro-Community Node (MCN)

In most cases the difference of configuration is not strictly in the basic hardware but rather in the controlling operating system and choice of installed peripherals and components. Readers are advised to note that the “Architecture” should not and cannot care about what is installed, but rather the Architecture employed should facilitate the services required of it. Therefore it is accurate to say that BMSTDA is hardware independent in nature. (Detailed configuration guidelines of each configuration is available from the author. These details have been omitted from this WIKI)

Obviously no single software application for BMSTDA devices exists today and there is an excellent opportunity for Ethiopian software developers to show their skills and develop complex, time-critical/mission-critical management software that provides:

  • Voice Switching
  • Packet Data Switching and Routing
  • Alarm Management and Disaster Mitigation
  • Remote Configuration and Troubleshooting
  • Automated Tasks ,li>Video/Audio Conference and Broadcast Services
  • Content Caching and Application Service Hosting
  • And many more….

The Micro-Community Communication Node is not a single device, and does not refer a collection of technology, but rather it is a moniker for a service network that provides applications and connectivity for an entire “small” area. The idea of Micro-Community Communications Node is a reference to the concept of “Community Communications Node” that acts as a tele-center in some villages in rural areas worldwide. It is assumed that a MCN would be setup in a small area serving several house-holds or established schools or medical facilities which would be entrusted with its safe-keeping and guardianship. In fact, that family or nearby families or nearby schools/hospitals can be provided with support so that they in turn can “own” their infrastructure after a period of time. This is called Community Ownership of BMSTDA network infrastructure. By adopting local ownership practices:

  • Local youths can be employed to service and maintain this equipment
  • Typically there will be 2 (two) persons per MCN who will be responsible for it
  • Equipment pilferage is drastically reduced

A single MCN must be connected to at least one other MCN for it to be part of a “network”. But since one of the major objectives of BMSTDA is to provide “fault-tolerant, partial-mesh” service each MCN that is connected should be connected at the earliest opportunity to another MCN somewhere nearby.

A single MCN therefore can consist of any number of:

  • BMSTDA:BR
  • BMSTDA:BWR
  • BMSTDA:MMRU
  • BMSTDA:MCBS
  • BMSTDA:MMTS
  • BMSTDA:IPTG
  • Internet Support Equipment: in order to distribute the connectivity, the network on the ground or in the air has to be spread out.

A “small BMSTDA service area” is defined as 100 sq. km of coverage which may include many MCN in a network.

A “medium BMSTDA service area” is defined as 500 sq. km of coverage which will include many MCN in a network.

A “large BMSTDA service area” is defined as 1000 sq. km of coverage which will include many MCN in a network.

BMSTDA: Why does it work so efficiently at low-cost?

BMSTDA concept is new, but its technology components are not. There has been no attempt to “discover” new product technology, new methods of sending data transmission, new ways to manufacture cheaper, faster, better. What is new is however the idea of using existing technical products in an innovative way, by deciding to define a set of parameters that will satisfy the average requirements of the poor region and then attempting to choose compatible products to fit them into the picture puzzle.

There are common sense formulas that have to be taken into account that actually makes the BMSTDA concept work, but those formulas are related to old, established scientific disciplines such as Radio Engineering, Mechanical Engineering and Computer Science. For example the idea that a small power signal can be sent large distances away is embodied in the concept of all communications theories and is calculated generally as a “link budget”, which can be used to mathematically prove or disprove a given communications link will work. Given that a typical Wireless LAN adapter puts out a maximum of 200 mW (this is special order, normal WLAN cards are 30 mW) it can be calculated from various standard formulas that the signal of that WLAN card if connected WITHOUT LOSS to a 24 dBi grid parabolic antenna then the signal will go as far as 30,000 meters if the antenna is fixed at 60 feet above ground level on both sides.

What data rate can be achieved in this type of ad-hoc point-to-point network depends upon the various Wireless LAN or Radio Adapters being used, and that specification can be calculated from the Link Budget formula which is a complex equation of:

  • Power at Transmitter Output Port in dBm units less
  • RF Feeder Cable Loss in dB units plus
  • Transmit Antenna Gain in dB units less
  • Free Space Loss in dB units less
  • Signal Loss due to atmospheric factors in dB units plus
  • Receive Antenna Gain in dB units less
  • RF Feeder Cable Loss in dB units less
  • Misc. coupling losses in dB units = Received Signal Strength (in dBm).

In order to deliver the maximum amount of power from one side of a network to other side, the signal needs to be sent with as much as “less” loss as possible; hence the idea of directly attaching Wireless LAN adapters and Radio Modems to antenna is of significant merit and importance to the concept of BMSTDA implementation.

By eliminating the need to have hundreds of feet of RF Coaxial cable from (perhaps) a BWR to an antenna expense is saved in implementation as well as gaining signal strength for longer range applications.

By adopting software-based approach to designing routing and multi-media systems and not adopting hardware based approach where product is custom designed for high reliability situations, flexibility of implementation is achieved. In case new features are required, or different configuration is required just a software upgrade can be implemented. In case that equipment has returned from service and needs to be retired, instead of retirement they can be recycled into other configurations. For example, MCBS and MMTS share hardware, but they also can be used in emergency as BWR, BR, MMRU and vice-versa. That way, a BMSTDA network manager can rest assured that at least the network will remain functional until repairs are made. Also, older generation of personal computers can be utilized by recycling their parts into custom-designed BWR, BR, MMRU, MCBS and MMTS with no limits. There are obviously many sources of recycled and obsolete parts of computers and peripherals and the items can be purchased at throw-away prices.

The elegance of the BMSTDA concept comes apparent if we look at a sample network diagram. This Diagram has been deleted from this Wiki, but is available on request from the author

This top-level network diagram shows an imaginary area of 4 Micro-Community Communication Nodes connected to each other. The mode of connection in this network is primarily through wireless links (could be Free-space Optics, pt-to-pt Microwave of either Spread Spectrum or Clean Carrier). A VSAT Earth Station is connected to a BWR in service area MCN#1, which can be treated as a gateway for the entire network. The functionality achieved in each MCN is listed as follows:

MCN#1: Satellite Connectivity, IP Telephony Services, Internet/Intranet Services, Video-Conferencing services and connections to MCN#3, MNC#4

MCN#2: IP Telephony Services, Video Conferencing Services, IP Telephone Service, and connection to MCN#4

MCN#3: Connectivity to MCN#1 and Micro-community Broadcast Station Service for TV/Radio transmission to surrounding community. In this node, there is no provision for computer connections locally.

MCN#4: Connectivity to MCN#2, MCN#1 with Internet Support Equipment serving a collection of PC users and IP Telephone Service.

The use of common equipment allows the designers of BMSTDA networks to make hybrid network topology possible, and the roll-out of the network can be implemented in stages as and when practical.

Each MCN has to be designed to be independent in makeup and function, and should be provided with adequate power (typically << 50 Watts, 250Volts AC) and the required supply could be provided adequately by Lead-Acid deep-discharge storage battery which would be charged by Mains supply, Solar Panels or AC/DC Generators if available. Otherwise if a battery-plant were to be setup and fully charged through an AC/DC generator, then the MCN could potentially be operating continuously for several weeks without additional charge.

As can be seen in the diagram, some nodes are connected only to another node, whereas there are nodes which are multiply connected. Typical network configuration for small-size networks will result in 1-4 connections per node. For long-distance networks 1-3 connections per node can be expected. If for some reason more than 3-4 connections are desired at any one particular MCN location, it is recommended to connect two or more MMRU or BWR back-to-back as can be seen has been done for the case of MCN#2 in the diagram.

For the first link for a MCN to connect to other MCN the designation of that link is “Redundant Radio Link – Primary” or RRLP. There is only one RRLP per MCN. Other links that emanate from that same MCN in any direction are called secondary links, so they would be designated as RRLS1, RRLS2 and so on.

It is very important to note that the links RRLn? actually refer in most cases to IP subnetworks and the total number RRLn? will be reflected in the number of active routes in the routing tables of the various BMSTDA devices. Since some form of dynamic routing protocol will be active at all times, any link that will be disconnected will be also deleted from the routing table, and any new link or MCN that comes on line will cause a change in the routing map – new routes can be added or subtracted or changed according to the network status at any time.

This dynamic routing feature is a strong benefit of BMSTDA networks and this can be amply demonstrated in the next diagram which was presented as early as 2000 calendar year.

Each MCN node is depicted here in this diagram as a BWR but in reality they can be any type, BWR, MMRU, BR, MCBS, MMTS etc. Since there are multiple paths between the various nodes, it is practical to consider whether and how the network will survive a catastrophic disaster that perhaps would disable or destroy (due to natural disaster) any number of nodes, such as (for example): MCN#3 and MCN#7.

As the reader can see, by tracking the remaining possible paths the network connectivity is left intact even missing the two nodes:

  • MCN1 is still connected to Nodes 4, 5
  • MCN2 is still connected to Nodes 9,10
  • MCN3 is inactive
  • MCN4 is still connected to Nodes 1,9,6
  • MCN5 is still connected to Nodes 8,1
  • MCN6 is still connected to Nodes 4,8
  • MCN7 is inactive
  • MCN8 is still connected to Nodes 6, 5
  • MCN9 is still connected to Nodes 10,4,2
  • MCN10 is still connected to Node 9,2

Note that “MTS” refers to Main Telecommunication Switch similar to a VSAT Earth Station or a regular Point-of-Presence of an established fixed public switched telephone network.