Wireless Mobile Phone Global Network: An In-depth View
By Mark McCarron (MarkMcCarron_ITT@hotmail.com)

Introduction

Well here it is the in-depth analysis of the new Wireless Mobile Phone Global network. In this article we will compare and analyze the competing wireless technologies available. Also included is the background to this systems development, a basic how-to guide to build the infrastructure, IP mobility, the finances behind it, the OS, services, handset features, connectivity, storage Devices, and security (including preventing packet fraud). As a bonus, I have also included the legal framework behind the system. Finally, in a section entitled 'The Ultimate Legal Precedent in US History', I clearly explain why the US telecommunications industry is unconstitutional and cite, no less than, 16 breaches of The First Amendment to the US Constitution.


Wireless technologies

There are several different types of wireless technology. We will discuss the two main contenders for this system the 54Mbit line-of-sight and the 3 GPP UMTS 16Mbit non-line of sight systems. We will begin with the 54Mbit line-of-sight, we will not go too in-depth into this as most will be familiar with the system.

Access points are used to create a central point of contact for all wireless connections within a given distance. Now in practical use a load of 30 connections per access point is the maximum for basic network services such as printing, file sharing and Internet browsing. When you need to stream multi-media, live video in our case, then the maximum connections to a single access point is around 10. Access Points automatically perform load balancing when configured on the same network. So, the more access points, the better throughput end users will receive and lower latency times on their connections. Access points can also be used to increase the network coverage. Now these limitations on connection load are affected by the type of router system involved as well as the reception capabilities. The router that will be used behind this project is an intelligent and self-organising system. Due to this, the packet management system streams are dealt with more effectively. Combined this with the layered protocol of FAST TCP, which can boost transmission by up to 6000%, then the implementation of this system will outstrip the rated capabilities.

Line-of-Sight communication is difficult to implement in an urban environment as the considerations are extensive. Careful planning and testing must be done at access point locations. This ensures a smooth installation, minimal packet loss, widest possible coverage, and good cross-communication with other access points for load balancing. Now I have seen a number of access points that have been modified to include a pre-amp. A pre-amp, also known as heat, boosts the signal and its range. This practice is illegal in most countries as it causes interference with other radio frequencies. A better solution is to boost the gain on the receiver only and use a high quality filter to remove the background noise. By boosting the gain on the receiver, very weak signals can be captured. In a wireless network where all receivers have their gain boosted, the net effect is almost as good as using a pre-amp and there is none of the associated cross interference. There are limits to how far the gain can be boosted on the receiver and is determined by the quality of the setup. The only thing that must be considered is that the 802.11a compliant equipment operates with only 4 different channels (8 with UNII1 and UNII2 combined and 12 in North America) and this could present a problem with close proximity of access points. Also, adjacent cells can have issues with adjacent channels due to sidebands. Therefore, complete backwards compatibility across the standard may in some cases have to be eliminated. Since certain major mobile phone operators only support certain bands, this appears not to be much of concern to them and, thus, it should not be a major concern for this project.

Another concern, at least in Europe, was the strict licensing of the 5GHz frequency. The UK just recently (12th Feb. 2003) deregulated this and this appears to be the case throughout Europe now, however, there maybe some regions (such as Spain) that will not allow it. The problem here with regulation is that radar and satellite belonging to NATO operated in that spectrum. To combat this, a system known as Dynamic Frequency Selection (DFS) was introduced that 'sensed' when these broadcasts were happening. Another feature added was Transmit Power Control (TPC) that controls the signal strength depending on how close it to an access point. The solution maybe to abandon the 802.11a 5GHz frequency and move the system to the 802.11g 2.4GHz frequency. The 802.11g standard is also compatible with 11Mbit 802.11b standard and has no regulation problems globally.

Airgo has announced a 108MB implementation and it increases the range between 2 to 6 times. As the current technology has a range of around 200m this means that this new system has an effective range of between 800m to 1.2Km. You can read about Airgo's new technology here at 'The Register' (http://www.theregister.co.uk/content/69/32380.html). Now Proxim had a 2x system that boosted the signal range and pushed the rate to 108Mbit but this boost was not allowed in Europe. Since Airgo is using a dual transmitter system (2 54Mbit chipsets), I cannot see how this would break European regulations as it is essentially the same principle as using two cards and load balancing between them, unless of course they are using some form of pre-amp that breaches regulations.

The next technology is from IP wireless, Time-division-duplex (TDD) is used, according to the 3GPP UMTS UTRA TD-CDMA standard using Release 99 of the standard. The system supports channelisation of 5 and 10 MHz (6 and 12 MHz in the MMDS band). The 3GPP standards include 3 alternative air-interface (physical layer) standards, which all share higher layer protocol stack and core network / back office architecture. These air-interface standards are FDD W-CDMA, High Chip Rate TD-CDMA, and low chip rate TD-SCDMA. TDD, includes a true N=1 reuse meaning that the Operator can deploy a network throughout a city with only 1 5 MHz RF channel for a 3.84 Mcps system and one 10 MHz channel for a 7.68 Mcps system.

This technology operates in a non line-of-sight environment, however, it does not have the penetration of the mobile phone network in terms of buildings, etc. The systems is deployed in cells that can cover up to 29KM. This systems operates at only 16Mbits in a combined downlink and uplink and is broadcast over the 2.5GHz spectrum. This type of technology requires licensing and the is a lengthy process. Also, it requires masts to be erected, however, it does not require truck rolls or smart antennas.

IP Wireless's technology appears to be slow for the needs of this project. The cell deployment is just moving in the same direction as the mobile phone operators. In terms of cost effectiveness, the deployment of the 54MB wireless system appears to be the simplest option. Also, the legislation requirements for IP wireless's implementation presents a major hurdle and trying to obtain them globally would be a nightmare.


The humble beginnings

If you have read this part before in another article, then you can move to the next section. I was investigating new transport technologies in Finland with a few friends and co-workers on the GIEIS project, about 20 in all. I was explaining to them the IrishWAN project, simply put, the project is an attempt to distribute a country-wide wireless Internet access for next to nothing. Several areas are up and running so far on pilot programs to iron out any mass deployment bugs. Each region is independent from each other, however, this is just for the test phase and will progress to a unified approach. We can currently deliver 11MB access to each user, as the network expands a fee of between £5 and £10 per month will most likely be introduced to fund the system and it will move to 54MB.

An idea came to us about using the system to massively undercut the telecommunications industry and resolving the 'last mile' issue. We had at our disposal 5 Nokia 3310's, 10 54MB wireless cards, 5 PCI and 5 PCMCIA. We quickly networked 5 machines and stuck a 54MB card in each and connected them over a 100MB ethernet network. This would simulate the backbone and router structure. Next we had to bypass the transmission circuits on the Nokia 3310 and interface the packet stream to the COM port on the laptops. We used the laptop and a modified script to translate the Nokia packet stream into a TCP/IP datagram and direct it to the 54MB PCMCIA card. The video was supplied by a standard webcam and the laptops were being used to simulate different phone features. Now the Nokia 3310 was not designed to display video, nor did it have the circuits to do it. So we used a second program on the laptop that merged and separated the video stream from the voice stream. The video stream was output to the laptops monitor and the voice stream to the Nokia 3310. We had to add a two second synchronisation slide to the video, so that we could manually synchronize the voice and video. We did not have the time to write an automatic system and this was just to demonstrate the principle. We also created a server application that would allow a conference call between all 5 systems, again we did not have the time to write components such as a packet loss mechanism. It took 36 hours.

When we tested the system there were some minor problems, however, we corrected them. We then placed a conference call and synchronised each of the video links. It was a complete success, there was some minor skipping due to packet loss but nothing that serious and a bit more software optimisation and better packet handling would resolve the issue. Also, a protocol like FAST TCP would eliminate the problem completely. With the pulsed-width modulation disabled, talk time was around 8 hours. We believe that by removing the pulsed-width transmission system and replacing it with a 54MB wireless transmission system we could increase average talk time to around 2 or 3 times the current length.

The technical considerations are pretty straight forward. The first issue was IP mobility and how this could be achieved, would we use some form of DHCP or static IP assignments. We concluded that the best method was to adopt the ethernet approach and hard-wire an IP number to a chip on board and have them assigned by the Internet authorities. So, the deployment of IPv6 may be short lived, as we may not have enough IP addressing space. I do not know the extent of adoption yet on IPv6, if possible we could just scrap it and move to IPv7m (m for mobile). Also, as this is a layered protocol system we could introduce a type of FLASH BIOS or EPROM chip that would allow updates to occur. Of course we would need to introduce some form of hardware lock, to stop virus' from flashing your mobile BIOS or any other chip.

The next issue with IP mobility is roaming, how does the system know where to deliver the packet? The current system relies on the fact that nodes are essentially static and hardwired. This allows ARP to resolve things such as the underlying ethernet or hardware address. The way to resolve this, again, is pretty straight forward. Every packet is sent to the host system from which the IP was allocated. Lets say it is IrishWAN. Each wireless router would know the location of the central server and inform it that it was in control of your signal. Now the controlling router would calculate your signal strength and compare it with the strength from each router within a 3Km radius and then forward your packets to the closest router to the signal until that router was confirmed by IrishWAN central server that it was now in control and packets were sent to it directly. Some form of Dynamic ARP will need to be developed and deployed at router level.

Bandwidth throttles may have to be put into some places were network demand is high. Your mobile will tell you how much bandwidth is available and light up symbols to tell you which services can be used. One-to-One two-way video can work on less than 1MB bandwidth and unless you need an entire conferencing service, you will never really notice any effect of low bandwidth availability. Internet access is provided by combining business, school and university Internet access country-wide over the intelligent and self-organising routers. After business hours these connections are wasted and it is also the peak time for home usage. As this service will be free to business for mobile use and free global roaming they should have not much objection to hosting a small router and some cabling.


The WAN Infrastructure and Finances

Right, let's get down to the bones of this system. According to Weismann Consulting (http://www.weisman-consultants.com/) that a mobile phone tower (mast) can cost anywhere between $75,000 to $2,000,000. For our purposes we will assume a basic cost of £75,000 (Sterling) for the cost of erecting a mast. For triangulation of signals at least three masts are needed and that would cost, at our assumed rate, £225,000 then add to the that the cost of the back-end infrastructure. A lot of money, I am sure we all agree on that and that is the heart of the problem. With the cost of this system, trying to push up the speeds means more investment and this is simply not profitable. We are going to propose here a system that is highly profitable and can replace the entire global telecommunications infrastructure.

Now, let's take IrishWAN as an example here. IrishWAN's most practical approach to infrastructure costs would be to introduce a fee of around £10 per month, or about £2.50 per week. This would result in a total annual bill of £120 per user. Now, let's take an outrageously low figure of only 4,000 subscribers in an average sized city. This would generate an annual turnover of £480,000. Now, we will forget about tax here for the moment, also as this is infrastructure is a legitimate business cost, we can write it off anyway and by spreading it across a number of years we can maximize that tax benefit.

Now for £280,000 we can purchase 400 routers of the intelligent and self-organising variety (£700 each). Essentially, with these routers all you do is plug them in and mount the aerial. That's it, the rest is automatic. Now to increase coverage, there is a nifty piece of hardware. It acts like a hub, but it is designed for aerials, 10 aerials (or more) can be attached to it and it senses which aerial is receiving and switches between them at a high rate buffering the data and forwarding it to the server. From the servers point-of-view, it is the same as a single aerial so no additional software or even drivers are required. This allows access points to be distributed over an incredibly large area and even inside pubs and nightclubs. These hubs are about £50 each and would provide over 4,000 access points for 400 routers totaling £20,000. As most aerials use normal co-ax cabling, at £0.20 per meter this means that 10Km can be purchased for £2,000. Each aerial can be bought for £5 under bulk purchasing conditions, a total cost of £20,000 for 4,000. So far, that is a total amount of £322,000. This can provide effective coverage for 6-8 sqKm.

Now we purchase 20 servers, these do not need to be high-end either, just good specification desktop PCs will do. Since all the software will be free and preconfigured (thanks to the Linux community), including security, all you do is load it in. 20 severs will cost about £20,000 and should include ethernet cards (100Mbit). Add to that a 2 16 port hubs, some patch leads, and a wireless router (or more if you need) and that will set you back about £1000. That is a total cost of £343,000, leaving £137,000.

Now two engineers can install 10 routers a day and that would take 40 days for complete installations. Give them another 40 days to complete each of the aerial installations. Add another 20 days to load and test the back-end servers. Then another 20 days to test the entire infrastructure and load testing. That's a total of 120 days or about 4 months. A good office space will cost around £20,000 per year in rent. Hire 1 person (£15,000) to input mobile phone IP addresses and direct debit details and leave out up to £20,000 for legal, accountancy and bank fees, etc. This leaves you with £82,000 or if the two engineers are running the company, a wage packet of £41,000 in your first year and no company debt. Completely breaking even within 12 months.

There will be other costs such as advertising, stationary, garbage collection and rates, etc. However, even if all these hidden costs amounted to massive £21,000. The two engineers would still make £30,000 each in their first year. By striking deals with manufacturers of paying for hardware over financial years great tax benefits can be reaped as well. Talk to your accountant.

Remember, this is for only 4,000 people. For a city of 200,000 people an annual turnover of £24 million is generated, assuming all adopt the system. With free phone calls, live-video, a £10 a month fee, free broadband access (for PCs as well as mobiles) that makes BT's broadband look like the postal service, who wouldn't?

Also, in business terms, this enterprise is literally a goldmine and the beginning of a new boom in the IT industry.


OS, Services, Handset features, Connectivity and Storage Devices

The OS that will run the system will be Linux. The distribution will be preconfigured by the best in the industry for rapid deployment to back-end systems. The distribution will also have features that will allow it to automatically integrate itself into an expanding network to minimise administration requirements. As every local back-end runs on the same principles, images of the distribution can be made ready for downloaded and loading. Also, being Linux, everyone can see the source code and know exactly what information is kept on them, eliminating any privacy concerns. In addition, updates can be spread across the world in minutes and mirrored at every location, even updates for mobile phones too.

The services provided to the handsets can now be free email, free SMS, free MMS, free Instant Messaging, free voice-mail, free video-mail, free voice calls, and free streaming real time video calls. Also, this helps the progression towards e-Government services such as doctor appointments, dentist appointments, etc. The system can also be supported by local advertising such as cinemas, theatres, and nightclubs. Just go to the menu on your phone and select 'What's on Tonight?', select your continent, select your country, and select your city and view the list.

One of the major new services will be webTV. Basically, everyone and their dog, can now open their own web based TV station and stream it over the web. You can watch your favourite shows on the move, play multiplayer games and a host of other activites.

These new handsets can really demonstrate the power of modern technology. Not only will they have the capability to connect to the Internet, but also, to each other either by using the main network in your country or directly to each other. No more running about trying to enter someones number, just point the two IR sensors (wireless would ask everyone in the neighbourhood) at each other and select number request, all the other person has to do is say yes or no. Its a great way to get someone's number in a bar, even if you are not talking to them or strike up a conversation using the instant messenger. Also, the handsets can act as broadband modems for your PC or laptop. With a click of a button you can change the profile to suit the network settings of the network you wish to connect to.

Combine this technology now with Sony's memory stick Pro series of 512MB and we have a serious piece of hardware. No more burning CD's or using floppy disks to transfer files. We could even give it a USB interface for computers without a wireless adapter. Other features of the new handsets is that they can now become a personal stereo too. Either listen to web radio or your favourite MP3's stored on the memory stick. With multiple memory sticks, you essentially have a removable harddrive with unlimited capacity. The handsets will run Linux and Linux will be the primary supported OS.

The possibilities are only limited by your imagination.


Security

The SSID is a unique identifier that identifies all Access Points and wireless clients as a member of your wireless network, clients with a different SSID can not gain access to the network via the Access Points. Globally, all SSID's will be the same to allow for roaming. Access rights are done at a higher level in tandem with the protocol stack.

WEP is a security protocol, specified in the IEEE Wireless Fidelity (Wi-Fi) 802.11b standard using a 40bit data encryption algorithm. WEP should be renamed the 'Waste-of-Space Encryption Protocol', even other variants of this such as 128bit are not suitable. So, we are going to introduce OpenPGP and use the Twofish cipher either in software form or on a FLASH BIOS. This will allow upgrades to be done. When an upgrade is available, your phone will tell you when you log onto the network and provide the option to download. This will then be stored on the memory stick and the phone will then ask to install the update. The OpenPGP hash will be check against the main servers and if everything is correct, it will then ask you to turn of your phone and disable the hardware lock. When the phone comes on again, it will not have network connectivity and the update will install. When finished, the phone will ask you to switch off and enable the hardware lock. If the hardware lock is disabled, the phone also disables the network connectivity for security reasons.

Now, authentication of a phone is performed by both the router and your wireless WAN provider. The IP address of your phone is registered in the database of the wireless WAN provider against your account details. So to authenticate at a router, you turn on your phone and your IP address is resolved and it is forwarded to the login system of your wireless WAN provider. The IP address is checked against the database and if a match occurs, and the account has an active status, then this server sends back a login request. You enter your username and password and submit. This is checked against the database and an random encrypted code is sent to the router (it enters its database) and the router forwards it to your phone. It is then immediately filtered to each router, that this router talks to, and they then do the same. The idea being that their transmission will be faster than any mode of transport available to you and by the time you get to the next router, your code will be there.

When you try to use any service, first your encrypted code is transmitted to the router. If it recognises the code, it then requests the phones IP address. It then broadcasts the IP address to your wireless LAN provider with a code request. When that code returns, the two codes are compared and only if they are the same does the router allow transmission to take place. Now to prevent a slightly complicated spoofing attack, if a second packet arrives with a different code, the session is terminated, details recorded and passed to the local servers attack logs. For this to work, each request would require a unique randomly reusable serial number.

After you login, every five seconds your IP is broadcast back to the main server by a pulse and this updates the server with the IP of the router you are closest to. Since there is no way to resolve the new IP address of the routers to geographic co-ordinates, no tracking is possible. If the Linux community wishes, we could have the routers perform a random IP swap with any of its neighbours at a random time. However, this maybe overkill and may make network troubleshooting a bit complicated, such as not being able to locate problem routers. In this case, a dynamic map would need to kept and that provides a tracking method.


The Ultimate Legal Precedent in US History

In releasing a system like this I was well aware that big business would essentially be able to move in and take control of the system. So, I had to devise a method of legally protecting the system and, if at all possible, make its introduction of the system a legal requirement. I went into deep study looking for legal technicalities that could be exercised to make this possible. After all, if murders can walk free from jail due to technicalities in law, then there is no reason that I can not use the same tactics to benefit everyone globally with free mobile communication. Then I found it, in the most unlikely place, The First Amendment to the US Constitution.

"Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances." — The First Amendment to the U.S. Constitution

Now you maybe wondering, how does this help. Well, let us first isolate the portions that are relevant:

"Congress shall make no law...prohibiting the free exercise thereof; or abridging the freedom of speech...or the right of the people peaceably to assemble"

Well the US dictionary, Merriam-Webster (http://www.m-w.com) has the following definitions listed under the words 'free' and 'Freedom'(The numbers are from the dictionary definition):

Free:
1 c : enjoying political independence or freedom from outside domination
3. b : not bound, confined, or detained by force
4 a : having no trade restrictions
5 a : having no obligations or commitments
7 a (1) : not obstructed or impeded
8 b : not confined to a particular position or place
10 : not costing or charging anything
11 a (1) : not united with, attached to, combined with, or mixed with something else 11 c : not permanently attached but able to move about
15 : open to all comers

Freedom:
1. a : the absence of necessity, coercion, or constraint in choice or action
1. h : unrestricted use

I am sure that the majority of you can now see what I am about to do.

I claim that charging for any form of telecommunications, within the US jurisdiction, whether it be voice or data, through any medium including, but not limited to, wireless, cabling, or satellite is a violation of The First Amendment to the US Constitution.

I will apply my reasoning in the context of The First Amendment to the US Constitution. It must be noted that under US law voice and text are considered forms of speech and the digital representation (data) is considered a form of speech also. Information is drawn from data, therefore, data must also be speech otherwise data could not be turned into information. Where you see '(Freedom 1h)' or something similar this refers to the dictionary definitions listed above, a 'breach' is something that does not agree with the accepted definition.

"US citizens do not enjoy the 'unrestricted use' (Freedom 1h) and the 'free exercise' of the 'freedom of speech' over telecommunications as the services are 'confined' (Free 3b) by charges imposed. This 'obligation or commitment' (breach of Free 5a) to 'pay' (breach of Free 10) is supported by the threat of 'detainment by force' (breach of Free 3b). In addition, it can be also said that the threat of 'detainment by force' (breach of Free 3b) due to the 'need to pay' (breach of Free 10) means that US citizens are being 'coerced' (breach of Freedom 1a) and have 'lost there rights to choice or action' (breach of Freedom 1a) when attempting 'free exercise' of their right to 'freedom of speech' over telecommunication systems. Telephone, Mobile, and Internet users are being 'restricted' (breach of Freedom 1h), 'obstructed and impeded' (breach of 7 a(1)) in their attempts of 'free exercise' of their right to 'freedom of speech'. The 'free exercise' of both the 'freedom of speech' and 'the right of the people peaceably to assemble', in relation to video conferencing, is being 'obstructed and impeded' (breach of Free 7 a(1)) by both the 'restricted usage' (breach of Freedom 1h) and 'applicable charges' (breach of Free 10) imposed. In relation to business, they are being obstructed, impeded (breach of Free 7a) and restricted in their trade (breach of Free 4a) when attempting 'free exercise' of their right to 'freedom of speech' by 'charges' (breach of Free 10). The 'charges' (breach of Free 10) are a direct violation of 'free exercise' and 'freedom of speech'. The metering of calls based upon time used is a direct violation of the 'freedom of speech', as the definition of freedom is unrestricted use. In clearer terms, imposing charges on telecommunications systems, prohibits the free exercise of the freedom of speech and the right of the people peaceably to assemble."

I think that should about cover it. I am sure this will be added to by the American people and move the telecommunications industry back into the hands of the people where it belongs. Now, to ensure that no big business can develop nor control the new system I would like to enact 15 Am Jur 2d: Charities (American Jurisprudence 2nd Ed. , Volume 15, Topic: Charities). This act, which I have read in full, essentially states that even with selfish motives a charity can be established as long as it does charitable work. So, I have worked out a clever way to use this, to essentially create a 'non-profit, profit making charity in a round-about way'. The principle goes something like this and is perfectly legal.

We setup a charity that is responsible for 'The Global Expansion of the Free Exercise of the Freedom of Speech'. A good and charitable cause. This is respected in International law under The Basic Human Right to Freedom and thus acceptable globally. This 'charity' will have a large number of members, each (or a team or board) who (that) is (are) responsible for at maximum 1 Million people within his/her (their) local area. Each person that wishes to exercise his/her right to 'unrestricted use' or 'freedom of speech' on this system must agree to assist the charity in the financial upkeep of the system to the value of $10 per month in the form of a donation. The members of the charity will not be in competition with each other, they are partners within the same charity. You may only be involved in one area. If your local area is unavailable then you have the right to expand the charity in an area not within your place of residence.

Now I bet you are wondering how does this 'profit-making in a round-about way' occur. Well, all sums are sent through the local representatives to the central charity. The central charity deducts 5% of the total amount received and holds that in a charitable fund to be distributed annually to good causes globally. These 'good causes' must have a direct impact on the quality of life, the expansion/enforcement of basic human rights and those in need. Now the sums left are distributed back from the central charity as 'running costs' to the representatives. From this, they are then expected to pay for wages, hardware investment, and general running costs as dictated by the standards set by the central charity. The rest of the money left, is their 'wages'. There is no limit imposed on how much your 'wages' can be. Also, the fund for good causes is expected to exceed $50 Billion annually.

Globally, all connections to this charity must be a 'overseas branch' of this charity otherwise peering arrangements will be denied. This is because big business could attempt to move in outside the US. The money in these cases is routed in the exact same manner. The idea behind this is that we can spread the global wealth more evenly whilst maintaining a solid economic infrastructure. Also, it means that we are not subject to the monopolies commission nor any form of anti-competitive legislation.

To begin, a website must be put in place to co-ordinate with the handset developers for standards compatibility, generating initial investment, develop guidelines and generally all associated issues. Once, this site is up and running we will pull together the hardware manufacturers and I will introduce this intelligent self-organising router and its creators.

Now that you all know, I will leave it to you.