URLLC (ultra-reliable low latency communications) is defined by 3GPP for 5G (NR) and aims to meet the extremely demanding requirements for latency and availability of services. 5G (NR) mobile networks supporting URLLC must provide low latency and minimize packet loss and out-of-order delivery.
I. URLLC Definition: ITU-R specifies a one-way user plane latency of 1 millisecond in 5G (NR) systems. This can be further defined by breaking down the URLLC acronym and analyzing its requirements:
• Ultra-high reliability requirements: Ranging from 99.99% for process monitoring to 99.999999% for industrial robots. This covers transmission packet loss and packet reordering – both of which need to be as low as possible.
• End-to-end low latency communication requirements: Application layer latency below 0.5-50 milliseconds, and 5G wireless interface latency below 1 millisecond.
II. URLLC Applications: Various application scenarios can fully utilize its ultra-reliable low latency, including:
Augmented reality/virtual reality and haptic interaction technologies allow users to experience artificially created realities or obtain additional information by overlaying real-world information. This technology has been applied in the entertainment industry, industrial applications such as warehouse management and field maintenance, and is expected to be applied in critical areas such as enhanced surgery.
As autonomous vehicles gradually replace human drivers, transportation will also benefit from URLLC. Vehicles and infrastructure utilize advanced sensors, artificial intelligence, and near-instantaneous communication technologies to significantly improve efficiency and safety. The main advantages of low latency are reflected in remote driving and sensor sharing.
Smart grids are improving power distribution, utilizing communication capabilities to achieve better power balance and detect and mitigate faults.
Motion control covers machine tools, printing, and packaging machinery. URLLC is expected to control the movement and rotating parts of machinery in a synchronized manner, thereby achieving high efficiency.
III. URLLC Standards
3GPP took the first step towards URLLC in its first 5G release, R15; its air interface was defined with a latency of 1 millisecond and a reliability of 99.999%. In NSA (Non-Standalone) network architecture, the core network and wireless signaling must rely on LTE, which cannot meet the end-to-end latency requirements of URLLC. 3GPP R16 defines the SA (Standalone) 5G architecture, which has an independent 5G core network and can operate without LTE, providing two important functions—network slicing and mobile edge computing (MEC).
IV. URLLC Driving Factors: End-to-end latency typically depends on network performance and the distance between the server and user equipment, both of which are optimized to accommodate URLLC applications, including:
4.1 Air Interface: Low latency optimization in 5G is achieved through flexible subcarrier spacing, scheduling optimized for low latency, and uplink grant-free transmission. Differential multiplexing, robust control channels, and HARQ enhancements are crucial for improving reliability.
With new subcarrier spacing, the subcarrier spacing can be adjusted from 15kHz to 240kHz. Larger spacing means shorter symbol duration, thus shortening the scheduling interval. The scheduling algorithm can schedule micro-timeslots, further reducing transmission latency. To avoid delays caused by requesting transmission resources, uplink grant-free transmission can be used.
Differential multiplexing uses multiple antennas at the receiver and transmitter to create independent spatial signal propagation paths, thus preventing single-link failures. To ensure reliability, NR aims to build robust control channels with low bit error rates; introducing new coding and using low modulation coding schemes (MCS) for transmission. The HARQ retransmission mechanism is enhanced by pre-allocating retransmission resources, thereby reducing latency and improving reliability.
4.2 Network Slicing: This is a key feature of 5G, allowing resources to be allocated on demand according to the service needs of different users. Resources are flexibly partitioned and isolated from the influence of other users, creating end-to-end logical channels. The required QoS for user slices can be configured on demand from the wireless interface to the core network. For example, for the same user, 5G can create a high-capacity video streaming slice for enhanced mobile broadband (eMBB) services without strict latency constraints; at the same time, it can also create a low-latency slice for ultra-reliable low-latency communication (URLLC) for robot control. Business Functionality - This feature is only applicable to the Standalone (SA) architecture of the 5G core network.
4.3 Mobile Edge Computing significantly reduces latency and improves reliability by hosting user applications on the "edge side" of the Cloud Radio Access Network (C-RAN). Therefore, transmission latency primarily depends on wireless access. Hosting at the edge avoids traversing the core network and reduces the number of nodes in the data path, thereby improving reliability.
