Iran's Internet Map: 7 Landing Stations, 6 Cables, and a 175 ms Trip to Next-Door Kuwait via Italy

Based on real measurements from GeoCables monitoring infrastructure, April 2026.
Iran sits on two warm-water coastlines — the Persian Gulf to the west and the Gulf of Oman to the southeast — facing some of the busiest submarine-cable corridors on the planet. Yet from a topology standpoint, the country is almost invisible to the global Internet backbone. None of the hyperscale trunks that pass within a few hundred kilometres of Iran's shore land there. To reach Iran's nearest neighbour, packets often travel through Frankfurt and Milan first. This article maps every cable that does touch Iran, and shows what the real measurements look like.

Seven landings, six cables

Iran has seven submarine-cable landing stations — the same count as the United Arab Emirates and one more than Saudi Arabia. The geographic distribution looks healthy on a map: three on the central Persian Gulf coast, two on the Gulf of Oman, and two offshore facilities in the Persian Gulf itself.

Counting landings is misleading here. What matters is which cables they connect to. Six unique cable systems touch Iranian soil, and every one of them is regional or bilateral.

The oldest cable here, UAE–Iran, was commissioned in 1992 — three years before the dot-com boom started, and well before sanctions made new submarine projects with Iran politically risky. The most recent, OMRAN/EPEG, dates to 2013. After that, nothing.

The bypass

The cables that don't land in Iran are more revealing than the ones that do. Look at any global trunk passing through the Persian Gulf or the Gulf of Oman:

• 2Africa (45,000 km, the longest cable in the world) — lands in UAE Kalba and Abu Dhabi, Oman Barka and Salalah, Saudi Yanbu and Jeddah, Kuwait City, Bahrain Manama, Iraq Al Faw. Skips Iran.
• AAE-1 (Asia–Africa–Europe-1) — lands in UAE Fujairah, Oman Al Bustan and Barka, Saudi Jeddah, Pakistan Karachi. Skips Iran.
• SEA-ME-WE 5 — lands in Oman Qalhat, Saudi Yanbu, UAE region, Pakistan Karachi. Skips Iran.
• SEA-ME-WE 6, EIG (Europe–India Gateway), India–Europe Xpress — same pattern.

The pattern is unambiguous. Hyperscale consortia routinely build cables whose physical path runs within 200 km of Bandar Abbas, but every one of them stops at Fujairah (UAE) or Salalah (Oman) instead. Of the 6 cables touching Iran, only FALCON connects to a global system — and FALCON is a 20-year-old private cable built by FLAG (now part of Reliance Communications) for its own customers, not a shared backbone.

The Sparkle detour: 175 ms to next-door Kuwait

The straight-line distance from Ganaveh on the Iranian coast to Kuwait City is roughly 250 km. The Kuwait–Iran submarine cable connecting them is 380 km long. Light through fibre travels that distance in about 1.9 ms one way, or 3.7 ms round trip. Theory says: any traffic between Iran and Kuwait should hit single-digit RTT.

Reality says: 175.87 ms. We pinged a Kuwait City endpoint (45.197.81.3, AS63139) from a probe in Ganaveh (probe ID 1006480) and traced the path. RIPE measurement 163280281:

The packet leaves Iran, climbs to Frankfurt over Sparkle's European backbone, hops to Milan, comes back south to Jeddah in Saudi Arabia, and finally lands in Kuwait. Total: seventeen hops and a 5,000 km detour to reach a city 250 km away. The Kuwait–Iran cable, built in 2005 specifically to connect these two coasts, carries no measurable IP traffic on this path.

The pattern repeats across every measurement we have. Three RIPE pings from Ganaveh to the same Kuwait endpoint at different times all returned 175.4–176.0 ms with the same Sparkle-routed path. Iran's chosen exit from the country isn't a Gulf neighbour — it's an Italian transit carrier.

The Western path: 244 ms to Los Angeles

Once a packet has travelled from Iran to Frankfurt, the rest of the world is cheap. From the same Ganaveh probe we measured Los Angeles (149.20.162.6, IO INC, AS21699). RIPE measurement 163797396:

Iran to LA: 12 hops, 244 ms. Iran to Kuwait: 17 hops, 176 ms. Distance to LA is 12,000 km; distance to Kuwait is 250 km. The two routes share the same European entry point (Frankfurt via Sparkle) and diverge only after that. From the Iranian network's perspective, Los Angeles is closer in hop count than Kuwait.

The other coast: 252 ms via Muscat and London

Iran's southern landings on the Gulf of Oman behave differently because they have working cables to Oman. Probe 65614 in Chabahar reaches Los Angeles via a more sensible-looking route. RIPE measurement 163764964:

Chabahar to Muscat is 6 ms — a real submarine measurement, almost certainly via the OMRAN/EPEG or POI cable, both of which were built in 2012–2013 specifically for Iran–Oman traffic. From Muscat, Zain Omantel carries the packet to London via overland fibre and onward over Zayo's North Atlantic backbone.

This is the rare case where an Iran cable does what it was designed for: carry traffic across the Gulf of Oman. The Persian Gulf cables (Kuwait–Iran in particular) do not.

The foreign domestic path: Jask to Chabahar via Dubai and Qatar

The most striking measurement in our Iran dataset is internal. Probe 6798 sits in Jask, on Iran's south coast. We pinged a host in Chabahar — also on Iran's south coast, about 700 km east of Jask. The two cities sit on the same country's coastline, on the same body of water, with two direct submarine cables (OMRAN/EPEG and POI Network) connecting them through Oman. RIPE measurement 163783562:

Iranian traffic from Iran to Iran goes via Dubai, then Qatar, then back to Iran. The probe sits in Jask but its first responding hop is in the United Arab Emirates — its upstream provider is foreign. The path then crosses the Gulf Bridge International network in Qatar before delivering the packet to Chabahar, 700 km from where it started but having traversed roughly 2,000 km of routed path.

This is what happens when a country's coastal landings have no shared domestic backbone. Each landing connects to whichever foreign carrier is willing to peer with it, and traffic between two Iranian cities on the same coast hops through two foreign countries to reach itself.

The offshore landings

Two of Iran's seven landing facilities are not on land at all. Khark Island (29.25°N, 50.31°E) is a 32 km² island 25 km off the coast, primarily known as Iran's main crude-oil export terminal. It hosts a landing for the Kuwait–Iran cable. Soroosh Platform (29.07°N, 49.48°E) is even more unusual: it's a fixed offshore oil-production platform in the Persian Gulf that doubles as a submarine-cable landing. The same Kuwait–Iran cable terminates there.

Hosting a cable on an oil platform is rare globally. Platforms are designed to handle hydrocarbons under pressure, not low-loss optical splicing. The arrangement on Soroosh predates most modern landings and reflects an era when Iran's oil infrastructure was the most reliable place to put new equipment offshore.

The 1992 cable

UAE–Iran, the 170 km link from Jask to the United Arab Emirates, has been in service for 33 years. It was built before the modern sanctions regime, before the Reliance/FLAG era, before the consortium model that defines today's cable industry. Its joint owners — the Telecommunication Infrastructure Company of Iran and Etisalat (now branded e&) — represent the only kind of partnership that still works in this part of the map: bilateral, between two state-controlled carriers, with no third-party investors to satisfy.

The cable has not been replaced or upgraded with a parallel system. Newer Iran cables (POI Network 2012, OMRAN/EPEG 2013) connect to Oman, not the UAE. The 1992 cable still carries whatever traffic Jask–UAE can support, alone.

Why the topology looks like this

Submarine cables get built when consortia of telecom operators and content companies decide a new route makes commercial sense. Modern consortia include US-headquartered hyperscalers — Microsoft, Meta, Google — whose participation is restricted by US sanctions on Iran. Even cables that don't include US owners face indirect pressure: equipment vendors (SubCom, ASN, NEC), financial institutions, and insurance markets all maintain their own compliance regimes. The result is that a cable consortium contemplating a Persian Gulf route finds it cheaper, faster, and less risky to land in UAE Fujairah (six global cables) or Saudi Yanbu (six global cables) than in any Iranian city.

Iran's response has been the opposite of greenfield cable construction: bilateral peering with whichever neighbour will accept the traffic, plus reliance on a small number of transit relationships with politically neutral carriers. Telecom Italia Sparkle (AS6762) — the Italian incumbent's wholesale arm — is the recurring name in our traces. Sparkle operates within EU sanctions but does not participate in US-side restrictions on Iranian transit, and its European backbone is the cheapest exit Iranian carriers have access to.

What the data shows, in three numbers

• 175 ms — Ganaveh to Kuwait City, 250 km apart, via Frankfurt and Milan. Theoretical minimum: 3.7 ms.
• 252 ms — Chabahar to Los Angeles, the only Iranian path that uses an Iranian cable for its first hop (to Oman). The remaining 246 ms is foreign infrastructure.
• 100 ms — Jask to Chabahar, two Iranian cities on the same coast, via the United Arab Emirates and Qatar.

Iran has the cables a country needs to be on the global Internet — seven landings, six systems, two coasts. What it doesn't have is the kind of backbone presence that turns those landings into a useful network. The cables exist, but the routing tables that would use them mostly don't.