Counter-drone jamming is being institutionalised as its target surface disappears

The incursion record

A maintained open-source ledger of Russian and Russia-attributed incursions counts 144 drone incursions over European airports and military sites since August 2024 ^[1]. Romania alone recorded at least 23 airspace violations in 2026, against 18 across the whole of 2022 to 2025 ^[1]. Baltic Air Policing has conducted three kinetic engagements of foreign drones since May 2026, the mission's first in 22 years ^[1]. On 30 July 2026 a Kh-101 cruise missile landed 100 kilometres inside Poland ^[1].

The political response has moved the other way. No ally has requested Article 4 consultations since Estonia's request of 19 September 2025, and engagement authority has shifted from national capitals into NATO's military chain through Combined Air Operations Centres ^[1]. Incidents are being normalised operationally at the same rate they are multiplying.

Leipzig/Halle is the case that should reset assumptions. On 4 August 2026 an FPV quadcopter carrying 800 grams of pentaerythritol tetranitrate struck the wing of an Antonov An-124, failed to detonate, and was found roughly four hours later by an airport bus driver ^[2]. The following day a Boeing 757 freighter collided with a drone at 400 metres while going around, and diverted to Hannover with nose damage ^[2]. Germany's federal interior minister described the attack as the work of professionals, probably on behalf of foreign powers; US officials told the Wall Street Journal that Russia likely orchestrated it; Russia's Berlin embassy denied responsibility on 7 August ^[2].

The recovered airframe carried dual SIM cards and a 5G router, permitting control from anywhere ^[2]. That detail matters more than the explosive. A drone commanded over a commercial mobile network is not defeated by jamming the hobby control bands.

What Europe is building

The European Commission published its Action Plan on Drone and Counter-Drone Security on 11 February 2026, organised around preparedness, detection capacity, coordinated response and defence readiness, with member states to appoint National Drone Security Coordinators and funding routed through Horizon Europe, the European Defence Fund, the European Defence Industry Programme and SAFE loans ^[3].

The European Drone Defence Initiative covers Poland, Romania, Estonia, Lithuania, Latvia, Czechia, Slovakia, Hungary and Bulgaria, with Eastern Flank Watch extending to Sweden and Norway under joint Finnish and Polish lead ^[4]. SAFE provides USD 175 billion in loans for defence acquisition ^[4]. EDDI targets operational status by 2030 and Eastern Flank Watch full functionality by end-2028 ^[4].

Its effector layers are kinetic systems, electronic warfare to sever control links and disrupt GNSS reception, directed energy, and cyber capabilities intended to recover an airframe intact ^[4]. Command-and-control architecture, sensor provisioning and interoperability between national systems remain unresolved ^[4]. Earlier instruments include EUR 250 million under the Border Management and Visa Instrument for border surveillance including counter-drone capability, and a European Competitiveness Fund line from 2028 ^[5]. The Commission has separately disbursed EUR 1 billion for Ukraine's drone capabilities under an EU-Ukraine Drone Deal ^[10].

Why the electromagnetic layer is eroding

Jamming defeats a drone by overwhelming its control link, and its effectiveness is a function of the jamming-to-signal ratio at the drone's receiver ^[8]. GNSS denial exploits how weak the satellite signal is at the surface, around -130 dBm; commercial jammers costing 40 to 80 US dollars deny navigation to receivers within hundreds of metres ^[8]. Both attacks require the target to depend on a radio signal.

Four developments remove that dependence ^[8]:

• Fibre-optic control. A hair-thin spooled cable eliminates radio emission entirely. By early 2026 more than 35 Ukrainian manufacturers produced fibre-optic FPV drones, and Russian adoption reached 30 to 50% in some units ^[8].
• Satellite links. Airframes fitted with Starlink terminals use upward-facing phased arrays that are difficult to jam from below ^[8].
• Autonomous navigation. Visual terrain matching, inertial measurement and onboard guidance need no external signal during the terminal phase ^[8].
• Mesh networking. Drones relay through multiple paths and route around interference ^[8].

Leipzig adds a fifth: cellular command and control. A drone on a public mobile network sits in licensed spectrum that a counter-UAS system beside a runway cannot lawfully or safely suppress ^[2].

The cost exchange is also less favourable than marginal-cost arguments suggest. Roughly 90 Russian jamming systems have been visually confirmed destroyed, damaged or captured in Ukraine ^[8]. A jammer is an emitter, and an emitter is a target.

Scale is the other problem. Kvertus proposed its Atlas concept as a 1,500-kilometre barrier of 8,500 detection and jamming units, with detection nodes identifying control and telemetry emissions up to 30 kilometres and jammers covering 0 to 6,000 MHz, seeking USD 130 million to complete deployment ^[11]. It is a substantial engineering answer that still assumes the target transmits.

The airport problem

Civil airports are where incursions bite hardest and where jamming is least usable. Copenhagen closed for close to four hours on 22 September 2025, affecting around 50 flights; Oslo saw multiple incidents within 48 hours; Billund, Aalborg and Esbjerg closed and Denmark imposed a nationwide civil drone ban; Munich suspended operations twice in 24 hours across 2 and 3 October 2025, affecting thousands of passengers ^[6].

No single method is comprehensive or risk-free in an airport environment ^[6]. Conventional radar is circumventable by smaller airframes ^[7]. Broad electromagnetic devices block an area to all drones rather than targeting one ^[7] — which beside a runway means degrading the navigation and communications the aircraft themselves depend on. Protocol-level takeover, landing an airframe without broad-spectrum jamming, avoids that collateral cost ^[8], but works only against control protocols it already understands.

What to do

• Characterise the control link, not the airframe. Fibre, cellular, satellite and fully autonomous profiles each defeat a different countermeasure, and only the first two are visible to RF detection at all.
• Buy detection that does not assume emission. Radar, acoustic and electro-optical sensing continue to work against a fibre-tethered or autonomous drone; RF detection fails silently and gives no indication it has failed.
• Settle spectrum authority before an incident. Any jamming near a civil airport degrades GNSS and communications for aircraft in the vicinity. That trade needs a named authority and a written threshold, not an improvised decision at 2am.
• Rehearse command and control. EDDI's unresolved C2 and interoperability architecture ^[4] is the layer most likely to fail first, and it is the cheapest to exercise now.
• Watch the diffusion of EW onto small platforms. BAE announced its Shadow EW family on 15 September 2026: three variants between 3.18 and 3.27 kg, receiver coverage from 70 MHz to 18 GHz depending on variant, under 150 to 250 watts, with production from 2027 to 2029 ^[9].

Domain view

Europe is standing up a counter-drone architecture whose effector layer leans on electronic attack, on a timeline running to 2028 and 2030 ^[4], against a threat that is shedding its radio dependence now. The mismatch is not that jamming stops working. It is that jamming is becoming one layer with a shrinking share of the problem, while procurement and doctrine are being written as though it were the main one.

The judgement worth making early is proportion: how much of a counter-UAS budget should follow the RF assumption, and how much should go to sensing and effectors that work against a target which never transmits.

For Malaysia and the wider region the incursion problem is European today, but the airframes are commercial, the techniques are published, and the airport vulnerability is universal. Two things are cheap to do before they are needed: detection that does not depend on the target cooperating by emitting, and a written answer to who may authorise jamming near an active runway.