The Invention of the Cellular Network
Signing in, signing out, handover
Handover during a call
Signaling and Error Correction
First Cellular Standards
Coopers „Cellular Call“
The Invention of the Cellular Network
The previous car phones were already very convenient. It was possible to make a call without the need for an operator. However, one major drawback remained. The capacity.
Every large city had practically only one base station which handled all phone calls within a radius of 50 to 100 km. Such a base station could have a limited number of radio channels. Thus, not many calls could be covered in parallel.
The solution of this problem was already researched in the same year in which the transistor was developed, 1947 and again it was at Bell Labs. The inventors of cellular radio as we know it today were two engineers named Douglas H. Ring and W. Rae Young. They first described the principle in a technical memorandum.
Instead of large base stations, a network of small radio cells is needed. Such a small radio cell has 6 neighbors if circular or hexagonal cells are assumed. All these neighbors cannot and should not operate on the same frequency in order to avoid interference. However, cells that are not direct neighbors can reuse the same frequencies. They are enough distant not to cause much interference. So it is required to spread a network of radio cells across the country and carefully plan the frequencies for this network. (Network Planning). The cells do not have to be of equal size. For example, in densely populated areas many small cells can be set up, while in rural areas large cells are sufficient.
Such scenarios were already considered in the sixties, but they failed because of the technology needed to manage such cells (more on this later). Furthermore, the construction of such a network was associated with high costs. As late as 1971, a consulting firm concluded that a cellular telephone network would have at most 100,000 users. Worldwide, there could be perhaps 200,000 users. Such assumptions led many countries and companies to shy away from such investments.
In the early 1980s, however, the pressure for more capacity was high enough to invest in a true cellular network. In the meantime, there were also special antennas that did not radiate in all directions (360°) like broadcast antennas, but only in one direction with an aperture angle of about 120°. This allows three antennas/base stations to be accommodated on one radio tower. This makes it possible to build a network with significantly fewer towers.
However, a cellular network comes with some problems. These are
- Centralized management of participants
- Transition between cells
- Signaling and handling of signaling errors.
Signing in, signing out, handover
In the pre-cellular era, a phone was active in a certain area. If someone wanted to call a mobile phone, he had to know where the subscriber was located.
To avoid this, it is required, that the cellular network had a central register that showed where (in which cell) the subscribers was located. Furthermore this register should indicate, if the mobile phone was active (ready to receive a call) or switched off.
Every cell phone that was switched on must therefore report to the central register and make its location known. This should of course be done automatically. Conversely, when a phone was switched off, it must sign out from the central register. This all was still relatively simple. It required especially a central computer with a database that was digitally connected to the mobile network.
A cellular network needs a central register which manages the subscriber and documents the location.
But what happens when the mobile phone moves, e.g. inside a car? It has to watch all the base stations in the area to see which base station is best for a connection. This choice can also be made by the base station but in any case both, the base station and the mobile phone must exchange information. If it is determined that the cell phone is leaving the range of a base station, the cell phone must be registered in a new cell or base station. The cell phone is thus passed on from one cell to the new cell. This is referred to as a handover. This means that a cell phone must be always active, even if it is not being used for a call. It transmits and receives all the time to observe its surrounding.
A cellular phone is constantly active even when no calls are being made.
Handover during a call
But what happens when a conversation takes place and the participant moves from one cell into a new cell at the same time? This can certainly happen in a moving car. The radio channels are now occupied by the voice transmission, but the radio environment must still be monitored and a possible handover must be performed. This task is done by introducing another channel next to the voice channel. This is located in the same channel. So there is a logical control channel and a logical voice channel. Both are in the same physical channel, (on the same frequency). This is called inband coding.
A good cellular system had to manage a handover by means of the control channels without disturbing the conversation.
Signaling and Error Correction
Due to the above-mentioned tasks within a cellular network, it is clear that more and more data must be transmitted between the base station and the mobile phone for control and monitoring. With radio transmission, however, it is almost unavoidable that errors occur during the transmission of data which could have fatal consequences. Already with the old B-Netz, information was therefore transmitted twice. When the data was received, it was at least possible to determine that an error had occurred. If this was the case, the transmission had to be repeated.
This back and forth is of course not efficient and in some also not really acceptable. It would be better not only to recognize errors but also to correct them. One possibility would be to transmit single information bits three times in a row and let „the majority“ decide. So if 1 is to be transmitted, one sends 111. If now e.g. 101 is received, it is obvious that the middle bit must be an error and can be corrected. But if two errors occur at the same time, e.g. 100, the error is corrected to a 0. This is not optimal and also the redundancy, i.e. the number of additional bits that have to be transmitted, is very high. However, there are better methods than this to correct errors. The following figure shows an error correction with 3 information bits and 4 redundancy bits.
The scheme uses three circles overlapping each other. Each circle has four sectors in which a bit can be positioned. The information bits are located between two circles. The redundancy bits are located in the circles and in the field where all three circles overlap. Let’s take 110 as the transmission value and write these values to the corresponding positions. The redundancy bits are now calculated in a way that there are always two 1 bits and two 0 bits per circle.
On reception, the rule is now checked with 2 times 0, and 2 times 1 per circle. If this is correct, the transmission was correct. However, if a bit is tilted, e.g. 100, you can see that the circle at the top left and the circle at the bottom have an error, while the circle at the top right is correct. Therefore the bit between the „wrong circles“ must be incorrect and can be corrected.
Parallel to the transmission or even storage of digital data, error correction algorithm were developed to detect and correct errors.
If a redundancy bit flips, this can also be detected and corrected. Only if two bits are transmitted incorrectly, an error is detected but cannot be corrected.
Error correction algorithms were developed in the sixties. This was done by mathematicians working on complex algebraic problems. An example for error correction are the BCH codes (Bose-Chaudhuri-Hocquenghem codes). These can correct many errors within long transmissions depending on redundancy.
A pioneer in the use of error correcting codes like the BCH was the Compact Disc which was developed by Philips and Sony at the end of the seventies. In this case, the error correction prevents a „crackle“ when playing a CD, for example, if it has a scratch.
First Cellular Standards
As with MTS and IMTS, Bell Laboratories was the leading institute in the development of a new standard for cellular systems. Although the cellular system had already been described in 1947, research on such systems was dormant for almost two decades. There was a lack of the necessary technology for its realization. However, starting in 1968, work resumed on a more precise specification of a cellular system. A team of three engineers Richard H. Frenkiel, Joel S. Engel and Philip T. Porter submitted a proposal in 1971. Further work resulted in a standard called Advanced Mobile Phone System (AMPS).
In 1975, AT&T obtained a license for an AMPS-based telephone network in Chicago. This went into operation in 1978. Motorola received a license for Baltimore-Washington. However, it took until 1983 for AMPS to be rolled out commercially in the USA in all cities and areas.
The first commercial cellular system was launched in Japan in 1979. Here, a standard called HCS, (High-Capacity System) was defined and later operated by the state-owned company NTT.
Shortly after NTT, the Scandinavian countries launched an advanced cellular system in 1981 with NMT (Nordic Mobile Telephone). Driven by the Swedish network operator Televerket, work on creating the NMT standard began as early as 1969. By 1975, the basic features of the standard had been defined, so that the development of the components could begin.
The UK simply adapted the AMPS system which was introduced as TACS in 1985. Germany went its own way and defined a completely independent standard of its own which went into operation in 1985 as the C-Netz.
Coopers „Cellular Call“
Today, the birth of mobile communications is often associated with a first call from Martin Cooper.
In the early 1970s, AT&T had a monopoly in the telephone network. It owned the long distance telephone business and also most of the local telephone operators. Additionally, AT&T was vertically integrated, meaning they not only operated all the networks but also built the infrastructure and devices. When Bell Laboratories defined the AMPS standard, the FCC was willing to grant AT&T the monopoly to operate a mobile communications.
This alarmed Motorola, which also had ambitions for mobile communications, especially because they had already built devices for MTS and IMTS. They tried as hard as they could to pretend another monopoly for AT&T in Communication.
In 1973, a Motorola team led by Martin Cooper built a state-of-the-art and portable handheld device to demonstrate a call to a landline. However, this was only a prototype and there was no network to make a “cellular call” yet. Instead, Motorola also built a base station that they connected to the landline network in New York. On April 3, 1973, Martin Cooper called a colleague at AT&T using the prototype and the base station. This was primarily done to impress the FCC, which ultimately succeeded. Motorola received a test license for AMPS in 1977 for the Baltimore-Washington area.
April 3, 1973 is today often considered to be the dawn of mobile communications, and Coopers call, the „first cellular call on earth“ which is actually complete nonsense. Just recently, on April 3, 2023, the 50th anniversary of this event was celebrated in the media.
