Providing adequate internet bandwidth to keep up with student and staff demand has been a problem at the TBI Bandung branches for some time. The Global English program has recently put additional demands on our service. TBI Jakarta has largely solved their internet woes with their current ADSL implementation, and we think it’s time that we addressed the problem here. Standard dialup modem solutions are proving to be slow and unreliable, no matter which ISP we use. (This has been no more apparent than during the past few days when our internet connection was downed completely by problems at TELKOM.) Additionally, with the expansion of the role of IT in both our curriculum and in facilitating our day-to-day administrative duties, the use of standard dialups is preventing us from exploring certain options that could make us more efficient and cost-effective—including the possibility of running our own mail and web servers, and of creating Virtual Private Networks (VPNs) to connect our disparate branches.

With these issues in mind, we propose to upgrade the internet services at TBI Dipatiukur as a preliminary to upgrading the services at both Bandung branches. We have explored three major alternatives to our current setup. These include wireless connections, ADSL, and leased lines. The latter solution, despite its benefits, is extremely cost-prohibitive at this time, so we will not be considering it here. That leaves wireless and ADSL.

A wireless solution would essentially mean running a cable from our servers in the Admin room up to an antenna on the roof. The signal is beamed to another antenna at the ISP a short distance down the road. From there it enters a high-speed digital line that routes to Jakarta, where it hits a satellite uplink and enters the Internet backbone. At no time are standard telephone lines used, so there is no risk of the kind of outage we have recently been experiencing, and there are no additional phone charges. Sending and receiving speeds are high-speed, symmetric, and the negative effects of poor weather are thought to be negligible. Monthly charges are reasonably low. However, the initial one-time outlay for equipment is not insignificant, due mostly to the necessity of building a transceiver tower on our premises.

The alternative is an ADSL (Asymmetric Digital Subscriber Line) solution, such as is currently being used in Jakarta. This would use a special modem and a standard phone line to connect to TELKOM, where the signal then enters a high-speed digital line and is routed as described in the wireless scenario above. Here, however, the sending and receiving speeds are asymmetric, i.e., the transmission speed is half that of the receiving speed. A 128kbps (kilobit per second) connection, therefore, would theoretically provide 128kbps downstream, and 64kbps upstream. The difference is not really noticeable until you start uploading files to a remote server—for example, when we upload changes to the TBI website or the Global English websites. An additional point to consider is that even though ADSL requires the use of a standard phone line, and will therefore incur local telephone charges, that line can still double as a regular telephone line at the same time it is being used for an internet connection. In terms of price, ADSL offers a much lower initial setup fee. However, the monthly charges are higher in comparison to wireless connections.

Here is a summary price comparison between these two solutions as provided by Melsa (complete breakdowns are available upon request). We have selected the 128kbps packages which, because they are shared connections, really provide an effective bandwidth of about 64kbps. Both options come with comparable extra feature packages, including 15 hrs/month normal dialup access for backup, and 14 real IP addresses (essential for running web/mail servers and for VPN tunneling).

You can see that if we were to choose the wireless option, after 3 months we would begin to save approximately Rp 4 million/month over the ADSL solution. Clearly, this is the more cost-effective route.

In addition to the lower cost, wireless offers several advantages that, in my estimation, make it a better choice for TBI.

• Overall bandwidth is greater (symmetric uplink/downlink speeds).
• We have freed up an additional phone line.
• We do not have to worry so much about problems at TELKOM
• If TBI-DU moves, or even if it is dissolved, we can still use the tower and other hardware at another location.

Since we are always interested in cutting costs on our monthly Internet bill without reducing our quality of service, I suggest that UniTech should join our network for the purpose of sharing our Internet connection. We could bill them at a similar (or even lower) rate than what they are currently paying. We reduce costs, and they end up with a faster Internet connection. Everyone wins.

Another possibility—and this is very interesting—would be for Jl. Jawa to use their existing two phone lines reserved for ISP connections to connect to our existing two reserved phone lines, and thereby share our wireless Internet connection. This would completely eliminate the need for a dialup ISP subscription at Jl. Jawa and likely give them a faster connection as well. They would still be paying local telephone rates, but the end result would still end up saving them money. We could test this only after a full wireless deployment at DU, of course, but it remains a distinct possibility.

In conclusion, I recommend we go wireless. Firstly, the large initial expense is quickly offset by lower monthly charges in comparison to ADSL. Secondly, it offers additional options to reduce costs. Finally, wireless appears to offer several technical advantages over ADSL, not the least of which are greater overall bandwidth and the elimination of our dependence on troublesome analog telephone lines.

Regards,

Matthew Arciniega
Microsoft Certified Systems Engineer
Senior Teacher for Information Technology, TBI Bandung
Email: matthew@tbi.co.id