Yesterday evening, during the dinner at a restaurant  at  Hawaii, I and my colleagues (Eric, Jun, Nihar and myself) along with a fellow colleague (Neycer) from Motorola were having some random ramblings. Somewhere in the course,came the topic on history of OFDM. It was indeed fascinating to trace the history. I did a bit of Googling later on and also traced some old notes from the discussion with Emre Telatar (who to me is a walking encylcopedia on several things). My information may not be too acurate, but roughly this is what I gathered after all the pile collection.

The origin of OFDM idea as such is largely attributed to Chang 1970.  Saltzberg had identified the problem of ISI and in came the notion of guard interval. Apparently, there is also a patent filed on this idea. The idea of cyclic prefix, the killer beauty which made OFDM ridiculously easy for equalization, was brought in by Peled and Ruiz in 1980. It was then Weinstein and Ebert who came up with the possibility of using FFT into OFDM. This traces back to the summer of 1971.

There are a few more interesting pre-work prior to these official OFDM milestones. Even though they are not really related, but hindlisht, we can still bring in similarities on how ideas shaped over time and different eras. For instance, the concept of parallel transmission was realized even in a product form in 1957 by a company Collins Radio Company. It was known as a Kineplex system. And the very idea of splitting to multiple carriers and power filling have signs of Gallager’s work and even waterfilling:-)

There is a Globecom paper which discusses all these. All these and may be more are neatly discussed there.