Telecommunication Network Reform: The Case for a Centralized Interconnection Framework in Bangladesh Voice Call Sector

Ahmed Ur Rahman (Romel):

As Bangladesh’s telecom sector stands at the crossroads of a new licensing regime, one critical question remains unanswered: should we reform what works or replace it without fully understanding the consequences? At the heart of this debate lies the role of the centralized interconnection framework—Interconnection Exchange (ICX), a mechanism that has ensured fair, efficient, and monitored voice service interconnection since 2007. Yet, the proposed regulatory reforms risk dismantling this proven system without offering a better alternative. Section 47 of the Bangladesh Telecommunication Act 2001 mandates a transparent and fair interconnection policy to prevent conflicts and discrimination among operators.

Despite this, the proposed reforms overlook these legal complexities in the name of liberalization, risking instability in the telecom ecosystem and disproportionately burdening smaller operators. Bilateral interconnection, as proposed in the new licensing framework, brings inherent and often irresolvable conflicts—particularly for smaller operators. For example, a small IP Telephony Service Provider (IPTSP) trying to join the market would need to connect with over 50 existing voice Access Network Service (ANS) operators, paying the full cost of interconnection. This makes the fiber cost too high to manage.

Conversely, if the regulator mandates existing operators to connect with new entrants to support SMEs, that’s not fair either. Why would an established operator absorb the cost of such interconnection when that yields negligible returns? This is exactly why a neutral intermediary interconnection like ICX is essential. Interestingly, even the new licensing draft seems to acknowledge this in Clause 7.2.5, which mandates that all ANSPs must establish interconnection or peering arrangements—either directly or through a third party. If such a third-party intermediary role remains necessary, then why not retain the ICX model that has already been serving this purpose effectively? ICX acts as a neutral hub for routing voice and SMS traffic between operators, simplifying interconnection between operators with a single point of physical connection. It lowers fixed costs, facilitates easier market access for smaller players, and—critically—offers regulators a valuable oversight on traffic monitoring and detecting anomalies.

Now, given the choice between two technically equivalent centralized systems —one with a built-in monitoring capability and one without— any prudent regulator will choose the one with greater oversight. However, government-owned BTCL, still outside the central hub, operates as a prime example of monopoly in interconnection, forcing all other operators to pay unnecessary interconnection cost. Figure: Voice network interconnection topology Yet, some critics have unfairly labeled ICX as a “middleman” that inflates call rates and other issues —an argument that does not hold up under scrutiny. The ICX service charge is 4 paisa per call, but ICXs keep less than 2 paisa after BTRC takes 50% and a yearly license fee of 1 crore taka. This minimal charge is fixed by the Commission under Section 47(2)(f) of the Telecom Act 2001, which mandates cost-based, profit-adjusted interconnection charges.

BTRC’s share acts as an extra markup for government revenue, which could be reconsidered in reforms. However, who benefits from it, as this charge does not impact consumer call rates? Since ICX launched in 2007, the fee has remained unchanged, while mobile operators repeatedly adjusted call rates and IPTSP and PSTN rates stayed stable for consumers under the same ecosystem. Thus, the belief that removing ICX will lower call costs is a fallacy, often perpetuated by misleading marketing narratives. ICX also doesn’t have any contribution to call drop and other quality issues, as it passes the call transparently to the next operator without alteration. Critics who claim that an additional interconnection node causes delay are also not supported by technical reality. Telecom networks are built to handle calls through several nodes, and all of this happens within just a fraction of milliseconds.

Critics also claim that ICX is a model unique to Bangladesh, but this is also far from the truth.

Many countries use third-party interconnection and exchange models where the operator density and fraud risk are high, where it is essential to prevent monopolies, keep competition fair, and encourage new entrants. One needs only to look at the recent shift in SMS traffic management in Bangladesh for proof of ICX’s effectiveness. Until June 2024, SMS interconnection was bilateral, and only six IPTSPs could access the market due to artificial resistance from dominant players. But since transitioning SMS routing to ICX in July 2024, over 20 IPTSPs have joined the ecosystem with minimal effort within months, enhancing both inclusivity and competition.

It’s surprising that the reform draft says that the existing architecture inhibits innovation and market entry. This conclusion appears to be based on limited consultation with the SMEs who have actually gained from ICX and ignoring the fact that ICX industry has recently spent over 100 crore taka for an ongoing technological migration that enables HD quality VoLTE inter-operators calls along with the support of all advanced voice technology i.e. VoWifi, Voice over 5G etc.

Ignoring the substantial contributions and domain expertise of those who have been at the forefront of systematically crafting a controlled and disciplined ecosystem over the years not only undermines their work, but also threatens the future security and integrity of Bangladesh’s telecom infrastructure.

In regulatory decision-making, the goal should never be disruption for the sake of novelty, but enhancement grounded in evidence. Especially in an era of rapid digital transformation, maintaining strong, centralized oversight is a necessity. Ultimately, telecom policy must prioritize sustainability, fairness, and national interest. Replacing the existing interconnection model with an unproven system would dismantle a working framework and harm innovation, competition, and consumer interests. The smarter path is not to discard ICX, but to modernize and integrate it further into the digital infrastructure. The choice should be simple: build on what works. Don’t fix what isn’t broken.

(The author is a Technology Expert, Head of Voice Operations, Agni Systems Ltd; Member Secretary, Technical and Legal Committee, Association of ICX Operators of Bangladesh)