FNB Network: routing economics and legal opacity

The ISP visible in the faint corporate shadow: FNB Network, routing economics and legal opacity in the Dhaka last-mile broadband market

The most interesting fact about FNB Network is not that it appears in Internet registry data. Many small ISPs do. The puzzle is that FNB is unusually readable as a network while remaining commercially opaque as a business. Its autonomous system, IPv4 block, IPv6 allocation, RPKI authorizations, tariff approval, public pricing pages, payment instructions, and address trail all indicate that a real operational network lurks behind the label. Yet these same pieces of evidence leave unresolved questions that matter for buyers, wholesalers, lenders, and competitors: who ultimately controls the operations, what is the formal renewal status after the last date on the public license list, whether the marketing talk of 'multi-upstream redundancy' matches actual routing redundancy, what the relationship hints with Coronet Corporation Limited actually mean, and to what extent the customer base is owned directly rather than reached through neighbourhood distribution.

This tension is at the heart of the intelligence problem. In legal due diligence, thin records often look like an absence. In Internet infrastructure analysis, however, small operators leave a different kind of audit trail: route entities, ASN records, contact details, tariff documents, payment portals, local address reuse, AS set membership, RPKI scope, and public service menus. These artifacts do not prove the same things as a commercial register extract or audited accounts. They prove that the operator has crossed several infrastructure legitimacy thresholds. They also expose the economic shape of the activity: a small Bangladeshi ISP on the periphery of Dakshinkhan/Azampur in Dhaka, selling regulated residential broadband and higher-ARPU business offerings, dependent on wholesale connectivity, constrained by IPv4 scarcity, and competing on installation, local trust, content access, billing, and support rather than on bandwidth alone.

The resulting thesis is simple: FNB Network appears as a genuine local broadband operator whose Internet numbering resources and public service pages reveal more about its economics than its corporate narrative. But the same routing evidence that validates operational presence also highlights concentration risk, governance frictions, and unresolved questions of ownership or counterparty. For a business and infrastructure reader, FNB is best understood not as a fully transparent company, nor as a mere directory entry, but as a neighbourhood access network that has partly institutionalized through APNIC resources, BTRC tariff approval, and visibility with the public ISP association.

From legal opacity to network visibility

The starting point is the RDAP/WHOIS trace. APNIC identifies ORG-FN18-AP as FNB Network, a LIR-type organization in Bangladesh, with an address at House #635/1, Hakim Mansion, Chalabon, Matir Mashjid Road, a phone number matching FNB's public contact trail, and the emailinfo@fnbnetworkbd.com. The APNIC organization record was last modified in September 2023. An RDAP-type record for AS149831 identifies the network as FNBNETWORK-AS-AP, managed by FNB Network in Bangladesh, with active status and a registration date in May 2022.

This is more than a mere commercial directory clue. An autonomous system number and portable IP resources are control points in the public Internet. They require registry interaction, contact details, maintainer entities, and routing authorization. They also create reputational and operational obligations: abuse contacts, route origin validation, acceptance by upstream providers, and, in practice, coordination with transit providers. FNB's public registry position thus makes it visible in a way that a small sole proprietor might not be visible in a corporate search alone.

But the legal picture of the business is thinner. FNB's website contains service pages, tariff references, plan menus, payment methods, and contact numbers, but its 'About' page is essentially under construction, not a meaningful corporate profile. Its footer and page text show an operational continuity language—including 'FNB Network' and, on a pricing page, a footer tag reading 'FNB Group'—but these labels do not establish a holding company structure. The public trail supports the existence of an operational label and a network organization. It does not answer, in itself, who owns the asset, how equity is financed, or whether the related labels are trademarks, historical names, billing aliases, or distinct legal vehicles.

This distinction is economically important. A customer buying a residential line at Tk500 may care primarily whether the connection works and whether a technician picks up the phone. A business customer buying a plan with real IP cares more about continuity, escalation, and contractual liability. A wholesale operator cares whether bills are paid and routes are clean. A buyer or lender cares whether the customer base, fiber points, AS number, bank accounts, and license can be transferred cleanly. FNB's evidence is strong on operational presence and weak on transferability.

Canonical identity: an operational label, several documentary shadows

The most defensible canonical identity is FNB Network, a Bangladesh-based ISP operating from the Dakshinkhan/Chalabon area in Dhaka and associated with AS number AS149831. This identity appears consistently in APNIC, BGP tools, the company website, tariff approval material, and ISP association lists. The APNIC organization record names FNB Network; BGP.tools identifies AS149831 as FNB Network; the BTRC tariff approval document is addressed to FNB Network; and ISPAB public search results identify an FNB Network member entry with the same license number visible on the BTRC license list.

The ambiguity begins at the edges. The address trail from APNIC and the company website points to House #635/1, Hakim Mansion, Chalabon, Dakshinkhan, Dhaka-1230. FNB's website repeats this address on its home and contact pages, with support numbers and the emailfnbnetwork2016@gmail.com. The BTRC's Upazila/Thana ISP license list and the ISPAB result, however, place FNB Network at House #693, Chalabon, Matir Masjid Road, Azampur, Dakshinkhan, Dhaka-1230. This may be banal: a move, a spelling inconsistency, neighbouring premises, a license address versus an operational office, or local address transliteration variance. But in a fragmented access market, address inconsistency is not trivial. It affects trust in the service area, notice delivery, license verification, bank KYC, and counterparties' ability to distinguish one neighbourhood operator from another.

There is also a naming ambiguity in the payment trail. FNB's payment page tells customers they can pay via bKash, Nagad, and bank accounts. It tells bKash users to find 'Friends Broadband Network' in the bill payment menu, while the same page lists a Dutch-Bangla Bank account under the personal name Md Shahadat Hossain and another Standard Bank account under FNB Network. This does not prove irregularity. In small ISP markets, founders' accounts, historical billing names, reseller names, and formal operational labels often coexist. But this creates a commercially relevant identity stack: the routing identity is FNB Network; the retail website is FNB Network; the bKash billing instruction uses Friends Broadband Network; and bank settlement appears to include both a personal and a business-named account. For an acquirer, auditor, or business buyer, this stack would require reconciliation before revenue could be treated as clearly attributable.

The social footprint reinforces operational identity but not the ownership story. A public Facebook page describes FNB Network BD as a BTRC-approved ISP and says it has provided Internet service since 2016; a related public group uses similar language and points to FNB's website. The website footer also carries a 2016–2024 continuity marker. This suggests the commercial service brand predates the allocation of the AS number by APNIC in 2022. The likely sequence is that FNB first operated as a local access provider or reseller, then gained greater network independence through its own AS number and IP resources. This sequence is common in access markets: a neighbourhood operator starts by reselling bandwidth under an upstream provider's numbering, builds a subscriber base and collection system, then formalizes routing and registry control once scale or credibility warrants it.

The public product is regulated broadband plus local service convenience

FNB's product pages are exceptionally useful because they reveal both the retail promise and margin constraints. The company markets home broadband, luxury plans, and business plans. The home page states that FNB provides cable/Ethernet Internet connections and lists monthly residential plans from 10 Mbps at Tk500 up to 60 Mbps at Tk2,500, with installation fees, no real IP, 'connected BDIX' language, gaming and social media features, and 24/7 support. The business page lists more expensive business plans from 15 Mbps at Tk2,000 up to 55 Mbps at Tk6,000, with lower business tiers including a 'real IP' and all tiers emphasizing duplex connectivity, support, BDIX, and claims of multiple upstream providers. A separate luxury page shows mid-to-high speed plans from 35 Mbps at Tk1,499 to 70 Mbps at Tk2,999.

The product logic is that of a local fixed broadband operator, not a pure hosting provider or backbone operator. The retail proposition combines bandwidth, local content, gaming latency, support, and neighbourhood availability. FNB's home page advertises unlimited usage, gaming, '18 hours double speed and 24 hours speed,' BTRC's 'Ek Desh Ek Rate' shared-line language, and 99% availability. It also lists links to movies, FTP, live TV, and software resources. The plan pages repeatedly mention BDIX connectivity and local content amenities, while the navigation includes private-address-style or local resource links and references to third-party torrent indexes.

These content and BDIX signals matter economically. In the Bangladesh fixed broadband market, advertised international bandwidth is not the whole customer experience. Users compare YouTube performance, Facebook reliability, gaming latency, local cache speed, FTP/media server availability, fault response, and whether the provider can quickly repair a distribution line. Low latency is important for gaming and video calls. A provider that can make popular traffic fast can defend a retail plan even when raw international transit is constrained.

The business product is a different margin instrument. A monthly plan of Tk2,000 to Tk6,000 has higher ARPU than a home plan of Tk500 to Tk800, but the buyer is more likely to demand availability, a real IP, an escalation process, and documentation. FNB's business tiers list a real IP for the 15 Mbps, 25 Mbps, and 35 Mbps plans, but the 45 Mbps and 55 Mbps business tiers state 'Real IP: No.' This pattern reveals IPv4 scarcity. FNB has 512 IPv4 addresses in its allocated /23; not all can be assigned to paying end users because routers, infrastructure, reserves, CGNAT design, and business customers consume space. A real public IPv4 address is therefore not a free feature. It is a scarce, monetizable input for SMEs, CCTV users, remote access needs, and small offices. The absence of a real IP on some higher-speed tiers is consistent with a provider rationing address space rather than bundling it automatically.

Revenue logic: prepaid discipline, low-ticket subscriptions, and margins heavily dependent on support

FNB's revenue model appears subscription-driven, with monthly broadband payments collected via mobile financial services, bank transfer, and an external billing portal. The website tells customers to pay the current month's Internet bill before the 15th and warns that the connection will be cut without notice by automated billing software if payment is not received. This billing rule is economically revealing. In low-ARPU access markets, collection discipline is not an administrative detail; it is the difference between positive and negative cash conversion. A Tk500 residential customer can quickly become unprofitable if support visits, payment delays, bKash fees, and wholesale bandwidth costs are not controlled.

The listed payment channels—bKash, Nagad, Dutch-Bangla Bank, Standard Bank, and an online billing portal—show that FNB tries to reduce collection frictions while retaining local flexibility. This combination is rational for a neighbourhood ISP. Doorstep cash collection scales poorly and creates leakage risk; purely digital collection may exclude some households; bank transfer is useful for business customers; mobile financial services are the natural middle ground. The external portal reference also suggests that billing is at least partially formalized through third-party software or white-label ISP billing infrastructure.

Gross margin pressure is structural. The company sells residential broadband at regulated mass-market prices. It must buy or lease upstream Internet capacity, domestic transport, NTTN or last-mile inputs, switches, OLTs, ONUs, routers, backup batteries, poles or building access, connection labour, and support. It must also absorb customer education, service calls, fibre cuts, payment follow-ups, and local disputes. Bandwidth costs may fall over time, but the support burden does not decline at the same pace. The economic question is therefore not just 'what is the price per Mbps?' It is 'how many paying accounts can be supported per field technician, per OLT port, per upstream commitment, and per collection agent?'

The product menu hints at the operator's response. FNB's entry-level plans follow BTRC's 'one country, one rate' structure, while the luxury and business plans create ARPU ladders. The company advertises free fiber optic connection promotions and 24/7 support. Free installation can accelerate subscriber acquisition but converts customer growth into upfront cash consumption. If churn is high, free installation destroys margin. If local trust and distribution line control keep churn low, free installation becomes a rational customer acquisition cost. FNB's model profitability therefore depends heavily on tenure: the same Tk500 subscriber is unattractive at two months and valuable at two years.

Switching costs are legally modest but operationally real. A household can change ISP, but it may require a new fibre line, ONU or router reconfiguration, a service visit, settlement of arrears, and coordination with building access or neighbourhood cable paths. The incumbent also benefits from habit and local trust: customers know which number to call, which technician covers the lane, and which payment reference works. FNB's public emphasis on phone, Messenger, WhatsApp, and 24/7 remote support reflects that the service channel is part of the product.

The ASN is small, real, and economically significant

AS149831 is the hard infrastructure anchor. BGP.tools lists FNB Network's AS number as registered in May 2022, active under APNIC, classified as access network, and originating two IPv4 prefixes: 103.187.124.0/24 and 103.187.125.0/24. It shows no IPv6 originating prefixes, one upstream provider and one peer, both visible as AS10075 Fiber@Home Global Limited in the IPv4 table.

This says several things at once. First, FNB is not merely a website brand or reseller without a routing footprint. It has its own AS number and originates address space. Second, it is small: two /24s equate to 512 IPv4 addresses. Third, it is an eyeball-type access network rather than a transit network; BGP.tools shows no significant downstream transit role for FNB. Fourth, its visible upstream dependence is concentrated. If Fiber@Home is the only globally observed upstream provider, FNB's Internet reachability materially depends on that relationship, even if some domestic, private, backup, or non-BGP arrangements exist behind the scenes.

The RPKI record reinforces the 'real network' conclusion. A route origin authorization authorizes AS149831 for 103.187.124.0/23, the component /24s, and an IPv6 block 2001:df0:bcc0::/48, with validation indicated as OK and validity extending until August 2026. RPKI is not a revenue statement, but it is a control signal. Operators that do not control or coordinate their numbering resources generally do not maintain valid ROAs for both IPv4 and IPv6. In practical terms, valid RPKI reduces the risk of route hijacking and makes FNB's prefixes more acceptable to upstream providers and networks enforcing route origin validation.

The IPv6 position is particularly interesting. APNIC data shows that FNB has allocated portable IPv6 space, and the RPKI record authorizes AS149831 to originate the /48. Yet BGP.tools shows no IPv6 prefixes currently originated by FNB. This gap is not uncommon among small ISPs, but it is commercially significant. IPv6 would reduce pressure on scarce IPv4 and improve long-term scalability, but deployment requires customer premises support, router configuration, support readiness, and upstream IPv6 transit. The unused IPv6 resource is a latent asset and a watchpoint: if FNB begins announcing IPv6, it would indicate operational maturation and could reduce reliance on public IPv4 scarcity as a product constraint.

There is a governance weakness in the registry data: the abuse contact records derived from APNIC include a note thatbidduth@fnbnetworkbd.comis invalid, with the abuse role last modified in January 2026. For consumers, this may be invisible. For upstream providers, security teams, and business customers, it matters. An invalid abuse contact can increase friction during abuse complaints, routing security incidents, blacklist disputes, and compliance reviews by upstream providers. It is a small administrative failure with disproportionate reputational leverage.

The upstream problem: marketing redundancy versus visible concentration

FNB's plans claim 'multi-upstream' connectivity. The BGP-visible picture is narrower: BGP.tools shows FNB with a single upstream provider, Fiber@Home Global Limited, and no visible IPv6 upstream path. This divergence should not be overinterpreted, but it should not be ignored.

Several possible explanations exist. FNB may have private domestic transport, BDIX access, or backup capacity that does not appear as a distinct global BGP upstream. It may receive Internet via Fiber@Home while using local caches, exchange fabrics, or reseller arrangements for domestic content. It may have had multiple upstreams at some point but not at the time of observation. Or the plan text may be generic marketing language copied across all plans. The commercially prudent interpretation is: the global Internet network dependency surface visible from public BGP is concentrated, regardless of whether the local access network has other operational redundancies.

Fiber@Home is a major connectivity provider in Bangladesh. BGP.tools identifies AS10075 as a Bangladeshi transport network with multiple upstreams, including Bharti Airtel, Tata Communications, Reliance Jio, Hurricane Electric, BSCCL, Lightstorm, and others; it also ranks Fiber@Home highly in Bangladesh in terms of AS cone and peer visibility. This makes Fiber@Home a credible upstream for a small ISP. It also means that FNB can outsource much of the international reachability complexity to a larger wholesale network. For a small operator, this is economically rational: a single strong upstream reduces procurement complexity and may be cheaper than building diverse transit relationships before scale warrants it.

The downside is bargaining power. A small access ISP with a single visible upstream has limited leverage over price, fault escalation, and route quality. If the upstream's economics change, if settlement terms tighten, if a compliance issue arises, or if FNB needs more peak-hour capacity, the operator has fewer credible alternatives. In the BTRC tariff economy, this matters because retail price flexibility is constrained while wholesale costs and QoS expectations remain in force. A regulated retail cap plus concentrated upstream dependency compresses margins.

The Coronet clue sits in this same wholesale layer. Public directory data associates FNB with Coronet Corp Ltd / Coronet Corporation Limited, but the public network evidence does not prove affiliation. Coronet's website presents the company as an IIG and IP transit provider in Bangladesh, offering DIA, IP Transit, MPLS, IPLC, and Global Ethernet services. BGP.tools identifies Coronet Corporation Limited as AS149765, an active Bangladeshi network with multiple upstreams and many peers/customers, and shows an AS set named AS149765:AS-CORONETIIG-BD. FNB's BGP.tools page lists membership in several AS sets, including Coronet's set, Fiber@Home-related sets, Summit-related, BSCCL-related sets, and others.

This is evidence of routing policy adjacency, not corporate control. AS set membership can reflect customer propagation, route filtering, legacy policy records, prior relationships, or third-party route management. It does not automatically mean Coronet owns FNB, operates FNB, bills FNB, or controls FNB's customers. The strongest inference is that FNB sits within a wholesale routing ecosystem where large IIG/transport networks manage or import customer routes. The weaker, unresolved hypothesis is that Coronet may be or have been a commercial counterparty. The evidence supports the former more than the latter.

Licensing and tariff economics in Bangladesh

FNB's public regulatory file is significant but requires careful reading. The BTRC tariff approval document dated July 2022 grants tariff approval in favour of FNB Network and reproduces the 'Ek Desh, Ek Rate' pricing table: 5 Mbps at Tk500, 10 Mbps at Tk800, 15 Mbps at Tk1,000, 20 Mbps at Tk1,200, 25 Mbps at Tk1,400, and 30 Mbps at Tk1,600, with a maximum shared contention ratio of 1:8. The same document sets out QoS and tariff conditions, including complaint record-keeping, bill relief related to downtime, and QoS requirements involving upstream redundancy, NTTN path redundancy, NOC capability, availability targets, and mean time to repair standards.

This framework changes the economics for a small ISP. In an unregulated market, an operator could freely widen the gap between advertised speed and actual service, raise prices in constrained neighbourhoods, or impose opaque degradation penalties. Under the BTRC tariff framework, at least on paper, retail price and contention are capped and QoS obligations are formalized. The operator's profit pool therefore comes from execution: better utilization, lower truck-roll cost, efficient upstream procurement, good churn control, local content routing, and sufficient customer density to spread fixed costs.

The public license list trail introduces sharper uncertainty. The BTRC Upazila/Thana ISP license list, dated as a public document in December 2024, includes FNB Network in Dakshinkhan with license number 14.32.0000.702.46.782.20.577, origination date 19 January 2022, and a license validity date indicated as 3 September 2024; the same line visually shows a next renewal date around early September 2024. The ISPAB public result for FNB Network shows the same BTRC license number and identifies the operator as holding an Upazila/Thana license.

The commercially important point is not that FNB lacks a license. The record supports a licensing history. The problem is that the public list available here does not by itself prove renewal beyond the September 2024 validity marker. Public lists may lag; renewal may have occurred outside the visible excerpt; or the operator may have continued while administrative renewal filings were not reflected in the reviewed version. But for due diligence, this is a hard watchpoint. A business customer, acquirer, or upstream provider would need to verify the current BTRC license status directly, not infer it from the website's 'BTRC-approved' language or historical tariff approval alone.

The same BTRC list also reveals local competitive density. Around the same Dakshinkhan area, the license list includes several other small ISP names, such as AK Network, AR Network, Arafat Siam BD.ISP, Diplomat Online BD, Samin's Network, and SMP Network, each with its own validity trail. This is not a monopolistic market structure. It is a fragmented local access environment in which neighbourhood operators compete alongside national operators, mobile broadband, and informal reseller channels. Fragmentation limits pricing power but can also protect incumbents through micro-geography: lane-level access, technician familiarity, building relationships, and existing connections matter.

Competition: the market is local even when the Internet is global

FNB competes in two markets at once. Upstream, it buys connectivity from large operators and sits within national and regional routing fabrics. Downstream, it competes street by street. The downstream market is not simply 'broadband in Bangladesh.' It is closer to 'fixed broadband for households and small offices reachable from FNB's local network in Dakshinkhan/Chalabon/Azampur.'

Substitutes are heterogeneous. A household may choose another local ISP, a cable/fibre reseller, mobile broadband, a national fixed operator when available, or a provider recommended by a neighbour. A small office may choose FNB, a larger ISP, mobile backup, DIA access from a larger provider, or a local reseller offering a cheaper but less formal plan. Each substitute attacks a different part of FNB's value proposition. Mobile broadband attacks installation friction and backup reliability, but may be weaker on unlimited fixed usage. Large ISPs attack credibility and scale, but may be less responsive at the lane level. Other neighbourhood ISPs attack price, installation speed, and personal support.

FNB's defense is therefore not just network engineering. It is integrated distribution. The company publishes multiple phone numbers, Messenger, WhatsApp, and Instagram contact options, and promises ticket response and 24/7 remote support. It markets gaming, BDIX, local content, and unlimited usage, which are precisely the features consumers feel without reading a service level agreement. In this market, a fast repair by a known technician can outweigh a theoretically superior upstream.

The economics of competition are also shaped by the BTRC tariff table. When entry-level plan prices are standardized or publicly constrained, providers compete on quality, freebies, support, and promotional bandwidth. FNB's home page language about free fibre optic connection and double speed periods must be read in this context. These are not random marketing flourishes. They are ways to create perceived surplus without formally breaching the tariff architecture.

But support-based differentiation has a cost. Every added customer increases potential fault tickets, router resets, payment questions, and outage complaints. Local ISPs are often caught between customers who experience broadband as a utility and an economy that resembles a small field-service business. If FNB exceeds its support capacity, churn and complaint costs rise. If it stays too small, it cannot effectively amortize upstream, staffing, and equipment investments. The profitable zone is dense, locally trusted, and operationally disciplined.

The ownership question: why unresolved control still matters

The public evidence does not currently establish a parent company or successor entity for FNB Network. The strongest canonical label remains FNB Network. The Coronet clue is best treated as a wholesale/routing policy or directory hint, not proof of ownership. The presence of both personal and business bank accounts on the payment page is a local commercialisation cue, not proof of legal structure. The 'FNB Group' footer is a branding clue, not proof of a corporate group.

This unresolved control question matters because the assets of an access ISP are bundled across multiple layers. FNB's value is not just its AS number. It is the customer book, the local network, the billing relationship, the field staff, neighbourhood permissions, routers, ONUs, upstream contracts, tariff approval, license status, bank settlement channels, brand reputation, and route entities. If these layers are owned or controlled by different people or entities, business risk increases. A buyer could acquire the brand but not the customer contracts; lease the network but not control the IP resources; obtain the customer book but inherit unresolved bank payee ambiguity; or rely on an upstream relationship that is personal rather than transferable.

The BTRC tariff document adds an important ownership clue without resolving the issue. The submitted approval copy is signed under an Owner, FNB Network label. An individual ownership structure would be consistent with the use of personal bank accounts and the formation of local ISPs. It would also mean that continuity, succession, and financing may depend heavily on the owner rather than a deeply institutionalized company. If FNB has since incorporated, reorganized, or transferred assets, that would materially alter the risk profile. The public artifacts examined here do not prove such a transition.

For lenders and wholesale counterparties, the practical due diligence questions are simple. Who signs the upstream contract? Who owns the last-mile network? Who controls the APNIC maintainer credentials? Who receives the bKash and bank settlements? What name appears on the BTRC renewal? Are customer contracts assignable? Are technicians employees, contractors, or reseller agents? The answers would alter recoverability, bargaining power, and business sales credibility. The public route trail validates the network, but it does not answer these legal control questions.

Competing hypotheses about FNB Network

The public evidence supports several competing hypotheses. They are not mutually exclusive, but differ in their commercial implications.

Hypothesis one: FNB is a locally originated ISP that matured from reseller to AS holder. This is the most likely reading. The social and website trail claims service continuity since 2016; the APNIC resources and AS number registration appear in 2022; BTRC tariff approval is also dated 2022; and the public plan pages resemble a genuine residential and SME broadband operator. Under this hypothesis, FNB first built a local customer base, then formalized as it grew. The economic implication is positive: local trust and existing connections can be valuable, while governance systems may still be founder-led.

Hypothesis two: FNB is operationally independent but commercially nested within larger wholesale and billing ecosystems. This is also plausible. FNB's visible upstream is Fiber@Home; BGP AS set membership points to broader routing policy ecosystems including Coronet, Summit, BSCCL, and Fiber@Home-related sets; the payment portal appears external; and the bKash billing instruction uses a different broadband name. Under this hypothesis, FNB owns the customer interface but depends heavily on external platforms and operators. The implication is mixed: low capex and fast deployment, but weaker bargaining power and greater counterparty exposure.

Hypothesis three: FNB's legal file is historically valid but administratively unresolved in the latest public list. The BTRC list shows a validity marker in September 2024, while the website and routing data remain active beyond that point. This may simply reflect list lag or renewal paperwork timing. But if the renewal were indeed unresolved, it would affect business sales, upstream compliance, and transaction value. The commercial implication is binary enough that current license verification should be considered mandatory.

Hypothesis four: Coronet is more than a routing policy clue. This remains unproven. Coronet is a real IIG/IP transit company with a visible network, multiple upstreams, and an AS set in which FNB appears to be included according to BGP.tools' AS set report. But no public evidence examined here establishes ownership, management control, or direct affiliation. If Coronet were a direct upstream provider, a reseller principal, or an acquisition counterparty, FNB's economics would change: it might have better wholesale pricing, stronger redundancy options, and a more institutional control layer. Without proof, this should remain a watchpoint rather than a conclusion.

What the evidence proves, suggests, and leaves unresolved

The evidence proves that FNB Network is not simply a directory artifact. It appears in APNIC organization records, holds an AS number, has IPv4 and IPv6 resources, maintains a valid RPKI authorization, originates two /24 IPv4s, is visible in BGP as an eyeball access network, and has public retail and business service pages. It also has a BTRC tariff approval history and appears in a BTRC Upazila/Thana ISP license list with a license number matching the ISPAB public result.

The evidence suggests that FNB is a small local fixed broadband ISP serving households and small businesses around Dakshinkhan/Chalabon/Azampur. Its economics are those of a dense last-mile operator: low monthly ARPU on the residential edge, higher ARPU for business customers, real IPv4 scarcity, dependence on wholesale transport, an operating model heavily reliant on support, and differentiation through local content, BDIX claims, gaming latency, and customer service.

The evidence also suggests a maturation path from informal or reseller-type service to formal network operation. The 2016 service continuity language on social and website pages predates the 2022 APNIC and tariff approval milestones. This is commercially plausible: local ISPs often acquire customers before acquiring independent numbering resources.

The unresolved facts are commercially consequential. The current license renewal status after the public list's September 2024 marker requires direct verification. The legal identity behind bank settlement and the bKash payee naming requires reconciliation. The actual redundancy architecture needs confirmation because public BGP shows only one upstream while plan pages claim multiple upstreams. The meaning of Coronet-related clues in terms of ownership or counterparty remains unproven. The invalid APNIC abuse email should be corrected or explained. IPv6 resources are authorized but not visible in routing, leaving unrealized scalability potential. Each of these uncertainties alters the risk premium applied to the business.

Evidence Register

Canonical operational identity and registry status. APNIC identifies ORG-FN18-AP as FNB Network, an LIR-type organization in Bangladesh, with address, phone, and email details matching the operator's public web footprint. This is high-confidence evidence that the operational label exists within the APNIC resource system. A separate RDAP-type ASN view identifies AS149831 as FNBNETWORK-AS-AP, active and based in Bangladesh.
AS number and routed resources. BGP.tools identifies AS149831 FNB Network as an active APNIC AS number registered in May 2022, classified as an access network. It shows two originated IPv4 prefixes, 103.187.124.0/24 and 103.187.125.0/24, and no visible originated IPv6 prefixes. This is high-confidence evidence of operational Internet visibility, but it does not reveal subscriber count, profitability, or ownership.
RPKI authorization. The RPKI record authorizes AS149831 for the IPv4 /23 and component /24s, plus the IPv6 2001:df0:bcc0::/48, with validation indicated as OK and validity extending until August 2026. This is high-confidence evidence of route origin control and resource hygiene. The commercial significance is that FNB's routed resource position is more mature than a purely informal reseller.
IPv6 gap. The APNIC and RPKI evidence shows allocated and authorized IPv6 resources, while BGP.tools shows no visible IPv6 origination. This is a high-confidence operational gap. It does not mean FNB lacks IPv6 resources; it means public routing shows no IPv6 deployment.
Upstream provider dependency. BGP.tools shows FNB with one visible upstream and one visible peer, both listed as AS10075 Fiber@Home Global Limited for IPv4. Fiber@Home itself is visible as a larger Bangladeshi transport network with multiple international and domestic upstream relationships. This is high-confidence public BGP evidence of concentration in globally visible routing, although it may not capture private domestic backup or local exchange arrangements.
Coronet relationship clue. Coronet Corporation Limited publicly presents itself as an IIG/IP transit provider in Bangladesh, and BGP.tools identifies Coronet as AS149765 with multiple upstreams and an AS set named AS149765:AS-CORONETIIG-BD. FNB's BGP.tools page lists membership in several AS sets, including Coronet's. This is medium-confidence evidence of routing policy adjacency or a wholesale market relationship. It is not proof of ownership or affiliation.
BTRC tariff approval. A BTRC tariff approval document dated July 2022 grants approval in favour of FNB Network and reproduces the 'Ek Desh, Ek Rate' plans, contention ratio, and QoS conditions. This is high-confidence evidence of regulatory interaction and tariff approval history. It should not be mistaken for a current license renewal certificate.
BTRC and ISPAB license list trail. The BTRC Upazila/Thana ISP license list includes FNB Network in Dakshinkhan with license number 14.32.0000.702.46.782.20.577, origination date 19 January 2022, and a validity/renewal marker in September 2024. The ISPAB public result for FNB Network shows the same license number and Upazila/Thana category. This is high-confidence evidence of licensing history, but not sufficient by itself to confirm current renewal beyond September 2024.
Public product and pricing pages. FNB's website lists home plans from Tk500 to Tk2,500, business plans from Tk2,000 to Tk6,000, and luxury plans from Tk1,499 to Tk2,999. The pages emphasize cable/Ethernet or FTTH-type service, BDIX connectivity, gaming, support, duplex connectivity, and, selectively, real IP for business tiers. This is high-confidence evidence of market positioning and revenue logic.
Payment and billing evidence. FNB's payment page tells customers to use bKash, Nagad, bank accounts, and an online portal, includes a disconnect-on-15th payment rule, and contains identity ambiguity through the 'Friends Broadband Network' bKash instructions and both personal and FNB-named bank accounts. This is high-confidence evidence of collection mechanisms and medium-confidence evidence of identity complexity. It does not prove ownership by a named individual.
Contact and support surface. The website publishes an office address in Dakshinkhan, support numbers, an email, Messenger, WhatsApp, and Instagram contact options, plus ticket response language. This is high-confidence evidence that FNB presents itself as an active, customer-facing ISP, not a passive AS number holder.
Social and continuity signals. Public Facebook search results identify FNB Network BD as a BTRC-approved ISP and indicate service continuity since 2016, with an associated group pointing to the same website. This is lower confidence than APNIC or BTRC evidence, but it supports the hypothesis that the retail service brand predates the 2022 ASN/tariff milestones.
Governance frictions. The abuse contact records derived from APNIC include a note that an FNB email address is invalid. This is high-confidence evidence of a registry contact issue. It is not evidence of network failure, but it is commercially relevant for upstream compliance, abuse handling, and business credibility.

Watchpoints

Current BTRC license renewal. The most important 12-month watchpoint is whether FNB's Upazila/Thana ISP license is publicly updated or directly verifiable beyond the September 2024 validity marker in the BTRC list. Historical approval and active routing are not a substitute for a current license in enterprise or transaction due diligence. A clean renewal would reduce regulatory risk; an unresolved renewal would increase the cost of capital, weaken business sales, and could affect upstream providers' willingness to carry routes.
Upstream diversification. Public BGP currently shows Fiber@Home as FNB's only visible upstream. A second visible upstream, particularly via Coronet, Summit, BSCCL, or another Bangladeshi carrier, would materially improve resilience and bargaining power. Continued single-upstream visibility would keep FNB economically dependent on a single wholesale relationship, even if local backup arrangements exist.
Coronet signal conversion to explicit. The Coronet clue should be watched for conversion from routing policy adjacency to explicit commercial evidence: customer lists, routing policy changes, invoices, press, website claims, PeeringDB updates, or downstream BGP visibility. If Coronet is only an AS set artifact, the business remains local and independent. If Coronet becomes a direct upstream, partner, acquirer, or operational principal, FNB's wholesale economics and institutional credibility change.
IPv6 announcement. FNB has IPv6 resources and an RPKI authorization, but no visible IPv6 route origination in the examined BGP source. IPv6 deployment would indicate network maturation, reduce long-term IPv4 pressure, and improve technical credibility. Continued non-deployment preserves the scarcity economics around public IPv4 but may limit future scalability.
Abuse contact correction. The invalid APNIC abuse contact note is a small governance defect, but visible. Correction would improve operational hygiene and reduce friction with carriers, security teams, and business customers. Non-correction increases the risk that abuse tickets, blacklists, or routing disputes become harder to resolve.
Real IP monetization and IPv4 scarcity. With only 512 IPv4 addresses, FNB's public IP policy is economically important. Watch whether real IP remains limited to selected business tiers, becomes a paid add-on, or is displaced by IPv6 and heavy CGNAT-based consumer service. Public IP scarcity can support SME ARPU, but it can also frustrate power users and increase support load.
Billing identity normalization. The coexistence of FNB Network, Friends Broadband Network, personal bank settlement, and FNB-named bank settlement should be watched. A cleaner payee identity would improve auditability and transaction value. Persistent ambiguity may be tolerable in retail broadband but problematic for enterprise purchases, financing, and M&A.
Address and brand consolidation. The difference between the House #635/1 and House #693 address trails should be reconciled in future public documents. Address consistency is not cosmetic; it affects license verification, customer trust, vendor contracts, and legal notices. A single, updated address across APNIC, BTRC, ISPAB, and the website would signal institutional housekeeping.
Local competitive churn in Dakshinkhan. The BTRC list shows multiple small operators in or around Dakshinkhan. FNB's economics will depend on its ability to defend dense customer clusters against neighbouring ISPs, mobile broadband, and large fixed players. Watch for price promotions, free installation offers, spikes in customer complaints, and visible neighbourhood expansion.
QoS enforcement. The BTRC tariff conditions include expectations for availability, downtime compensation, complaint keeping, and redundancy. Stricter enforcement would favour operators with better NOC processes and redundant upstream/NTTN paths. Weak enforcement would preserve the current local-trust model but sustain uneven quality. FNB's claimed 99% availability and the BTRC QoS obligations should be tested against outage traces and customer complaints over time.
Website and portal modernization. A more complete corporate website, an updated About page, current tariff approvals, a clearer legal name, and an integrated payment portal would indicate institutionalization. Persistent under-construction corporate pages alongside active service pages would reinforce the current pattern: operationally real, commercially useful, but legally thin.
Route origin and prefix stability. FNB's two /24s and valid ROAs should be monitored for withdrawals, invalid RPKI states, origin changes, new prefixes, or deaggregation. Stable route origin behaviour supports continuity. Route instability, unexplained origin changes, or expired ROAs would be early warning signals of operational stress, upstream conflicts, or control changes.