Before the Dawn of IP Telephony - Part 21Difficulty in finding a location for testing PHS (March ~ July 2000)
These contents translated a serialization article carried by ITPro IP telephony ONLINE published by Nikkei Business Publications, Inc. Jump to the original (Japanese).
Shinji Usuba
General Manager
eSound Venture Unit
Oki Electric Industry Co., Ltd
We started the development of the second generation IP-PBX. Our challenge this time was PHS. Handling PHS was extremely difficult for IP networks that result in communication error when synchronization is not properly made.
First, we attempted to verify the algorithm using actual devices. As explained previously, accommodating PHS is basically the same as guaranteeing synchronous communication on an IP network. It was an extremely difficult technical task. Verifications of the actual device helped play a critical role of establishing this technology. We repeated endlessly the steps of finding algorithm error, fixing the error and performing verification again using actual devices.
When developing a new product by applying a basic method that has already been established in an existing product, verification items tend to shift very little from what is first expected. However, verifications from the basic method require more time effort than one would expect. This verification was no different.
Some could not understand the need to incorporate a system for synchronization on LAN. But in order to realize the basic concept of integrating intra-office communication on LAN and IP networks, we did not have a choice. With such thoughts in our minds, we repeated the verification with much time and labor. And as a result, the level of the algorithm for synchronization in an asynchronous network was gradually reaching perfection. However, there was yet another task we had to overcome to accommodate full PHS support.
Difficulty in finding a location for PHS verification
The algorithm was near completion and the system for synchronizing PHS communication among units placed on a LAN was finished. However, the system of synchronization takes some time before it becomes stable. Hence, completely identical operation as a conventional line exchange network could not be guaranteed. Since time is required for the system to become stable, we needed to verify how such time would influence the actual operation of PHS. Since we had performed development of a PHS system for offices at OKI in the past, I first thought that we could easily perform verification using the same development environment.
Photo 1 Shield room
However, an environment such as the shield room (photo 1) often used for design verification was insufficient for such assessment. We could not assess the borderline performance due to the short distance between base stations. Then, we thought of performing verification outdoors. But it was not easy establishing a test environment with the widespread deployment of public PHS with 1.9GHz band waves flying in every direction no matter where we tried. We couldn't find an appropriate location for verifying a private branch PHS. Although I knew verification of the synchronization system was necessary, searching for the location was harder than it seemed. Members starting with Koji who was organizing the development were distressed. It was the flexible imagination of K, in his third year at OKICOMTEC, who found the solution.
Verification in a basement cafeteria
K found a location where there are no outside electrical waves and where sufficient distance can be secured. To my surprise, this ideal location was the basement cafeteria of OKI. At the time, the in-house cafeteria was located in the basement to accommodate the several thousand employees working at Shibaura, Tokyo during lunch. Although the cafeteria was used during the day, it was unused during the night. The cafeteria would surely compensate for the distance that could not be provided in the shield room. Young K made us aware that the perfect location was in fact, directly beneath our feet.
The idea was adopted immediately by securing the location, and verification for accommodating PHS on LAN started. Obviously, we were not able to use the place during the day as it operated as a cafeteria. So, we repeated the process of installing and testing the devices during the night and cleaning up during the day. Creating an environment needed for verification was time consuming and required much patience. As a result, we acquired important technical experience for accommodating PHS on LAN. I believe that this knowledge will become important for realizing rich communication in full IP and full broadband networks of the future.
Corporate VoIP completed
The full IP-PBX system was released into the market in August 2000. The accommodation of PHS realized the basic concept of guaranteeing communication regardless of the signal attribute including faxes and modems for data flowing through the network. Furthermore, the concept development announced at ISS '97 was almost finished.
The development of a corporate solution as VoIP gateway seemed to come to a rest with the accommodation of PHS in IP stage. PCs were becoming common within offices, port unit cost of LAN switches dropped drastically as we expected and the advantages of integrating infrastructures were slowly becoming known. Although this product was technically innovative, there were still a number of realistic issues remaining. Despite the understanding of the potential in the future, we needed to overcome various issues before determining a price, and specifications, and having the device used in the environments of our customers.
The shift to IP was surely happening, it was a long road before the technology was standardized as IP telephony and gained acceptance. We knew that voice IP packets would spread sooner or later by steadily evolving corporate solutions, but we believed it was something that would not happen very soon.
However, we are hit by a catastrophic event before a year had passed since the PHS accommodation in IP-PBX.
... To be continued
