How Quantum computing will disrupt the telco sector?

Introduction to quantum computers

Computation relies on bits, which can be in one of two states called 0 and 1. Operations which manipulate bits – flipping 0 to 1, for instance – are called logic gates. Computers perform tasks by combining bit inputs and logic gates.

Quantum computers use superposition to replace bits with qubits. A qubit’s state has a component corresponding to 0, and a component corresponding to 1, and carries information about both. To illustrate the impact, consider that two bits can be in one of 22 = 4 combinations. Four strings of two bits are needed to describe all combinations – but a quantum computer only needs one string of two qubits, which are in a superposition of all four combinations.

This impact scales dramatically: one string of ten qubits can represent a superposition of up to 210 = 1024 bit strings, and one string of 20 qubits can represent a superposition of up to 220 = 1 048 576 bit strings. This means quantum algorithms can potentially work exponentially faster than their non-quantum counterparts, providing a significant advantage for performing computationally expensive processes – some examples of which are given in the next section.

Quantum computers with enough qubits to carry out useful calculations are still several years away – whilst maximum capacity has grown from 65 qubits in 2020, to a new record of 1225 in 2023, IBM aims to build an error-correcting quantum computer, which is protected from noise and loss of superposition, with a capacity of 100 000-qubits (i.e. a further increase in qubits by more than 80x) by 2033.

Whilst we are waiting for this, the current era is known as NISQ: noisy intermediate scale quantum. These quantum computers typically have fewer than 1000 qubits and require challenging output processing to mitigate the effect of noise and loss of superposition.

Implications for telecoms

Cryptography

Telco networks carry sensitive information such as proprietary data, subscriber payment details, and government information. This is kept safe by encryptions like RSA-2048, which would take approximately 300 trillion years to break using classical computing. However, a sufficiently large quantum computer would take just 100 days, placing network data at risk. Telcos need to take action to protect their networks.

Two main approaches for quantum-safe encryption exist:

• Quantum key distribution (QKD): QKD involves sending private cryptographic keys along quantum channels. Observation of a quantum system unavoidably changes its state, making it impossible to eavesdrop on a quantum channel without detection.
• Post-quantum cryptography: this requires public-key standards which are unfeasible for even quantum computers to break. The USA’s standards body, NIST, is running an ongoing process to select the most promising candidates. Current forerunners utilise mathematical properties of elliptic curves and lattices.

Telcos around the world are already investing in quantum-safe networking using these methods. South Korea’s SK Telecom has developed a QKD-based VPN, the Madrid Quantum Communication Infrastructure project functions as a testbed for the coexistence of quantum and classical networks, and in 2022 the GSMA launched the Post-Quantum Telco Network Taskforce along with Vodafone and IBM. Furthermore, the European Commission recently published a recommendation for Member States to design quantum-safe roadmaps ‘as soon as possible’.

Network optimisation

Telco networks are extremely complex: different ‘nodes’ such as sites, subscribers, equipment, and even geography interact to generate a vast quantity of data. Operators must optimise deployments to maximise downlink speeds, minimise energy consumption, or meet network coverage targets. However, such analyses can quickly reach computational limits, with calculations that may take days to reach an approximate answer.

Traditionally, this complexity is handled by creating models which simplify the calculations through in-depth network expertise and dimensioning rules, thereby reducing computational load to arrive at a ‘good-enough’ result in a much shorter period of time. This approach can be highly effective in many situations, especially when high-level strategic decisions need to be taken. For example, Aetha’s own mobile network models dimension for coverage and capacity requirements over a 10–20-year horizon within a matter of seconds.

However, the increase in computational capabilities from quantum computing may offer an alternative route, and a potential integration into daily operational routines. As a (highly technical) example, so-called ‘quantum annealing, an effective NISQ technique, relies on the fact that physical systems tend towards the lowest energy state. By mapping an optimisation process to a quantum system which evolves under controlled conditions, you can find the optimised solution by identifying the lowest energy state of the optimisation process.

Work is already progressing on introducing quantum annealing across a range of network design problems. Examples include optimising power gains in MIMO vector perturbation precoding, increased efficiency in 5G vRAN network scheduling problems, and optimised PCI assignment.

The expected outcome of all these processes is a significant improvement in network operation, which in turn translates into higher network performance and, ideally, lower network opex and capex. The increased efficiency gains granted by quantum optimisation could prove invaluable for telcos to meet the data demands of 6G while minimising energy use.

What’s next?

While fully matured quantum computers are still several years away, understanding the threats and opportunities they pose will enable telcos to prepare for their arrival. Telcos should act now to prepare for the quantum cryptographic threat – an endeavour already taking place around the world, and with increasing support from policymakers and awareness from industry bodies.

At the same time, networks are moving towards 6G, and an ever-more data-intensive world, increasing the computational strain placed on existing operational systems. The opportunities afforded by quantum computation may be significant and use cases in network design – from mMIMO precoding to PCI assignment – are already emerging. As an industry, it will be important to think critically in the coming years about how and where new approaches can be best be deployed to realise powerful network efficiency gains – and quantum computing may ultimately be one of the tools to realise this.

Author