Skip navigation
Optimizing shared networks for reliable communications during major events

Optimizing shared networks for reliable communications during major events

High-load handling during emergency situations in large scale events

Large-scale events pose unique communication challenges, requiring reliable networks for attendees and public safety personnel. RAN sharing offers an effective solution, enabling public safety operators to utilize commercial networks while ensuring dedicated services. Discover how Ericsson's innovative approaches enhance connectivity and safety during critical events.

Introduction

Large scale events held in indoor or outdoor venues (e.g., stadiums, arenas, theaters, conference centers, etc.) are characterized by large variations in the number of users. These users do not only include the event attendees but also the organizers, public safety users (e.g., police officers, firefighters, first aid personnel, etc.) and even people passing by, near the event area. Compared to the usual daily traffic, during an event, high peaks in network usage can occur. During extremes, it is common for the users to experience some sort of service degradation (e.g., slower throughput, difficulties accessing the network and voice quality) throughout these high peaks. Service degradation affects all users but, whereas an event attendee may tolerate a slow connection while uploading a picture, there is a safety risk if public safety personnel cannot communicate. Therefore, it is important to consider the different requirements from the different user groups even during drastic fluctuations in user traffic.

Due to their importance in society, governments have been investing in public safety networks to provide communication services to public safety users, or anyone involved in public protection. However, deploying a parallel network only dedicated to public safety may be unaffordable or even unfeasible in some cases. Therefore, an increasingly common approach is for the public safety operator to deploy its own dedicated core network while sharing the radio access network (RAN) from a commercial operator, which can be done by using either MOCN (Multi-Operator Core Network) or MORAN (Multi-Operator Radio Access Network) architectures. This helps to speed up the network deployment of advanced critical broadband services whilst lowering the overall network costs.

The concept of RAN sharing allows the public safety and commercial operators to share either partially or entirely the use of one or more RAN sites that include the baseband*, the radio and spectrum. Also, RAN sharing features are essential to appropriately distribute the shared resources between public users and public safety users, ensuring quality functionality and operations.

Ericsson offers various deployment options for public safety networks, one of which involves a dedicated core and shared RAN network, as previously described. To better understand the implementation and operational challenges faced by major public safety operators, Ericsson has collaborated closely with them to understand the issues and develop appropriate solutions. Through this collaboration, it was noted the occurrence of the congestion issue mentioned above. Subsequent analysis led to the identification and testing of corrective measures, the findings of which are outlined in the following sections.

Download paper

* Baseband is the equipment that implements the function of the eNB in the case of an LTE (Long Term Evolution) network or a gNB in the case of an NR (New Radio) network.

Potential service degradation for public safety users during high network loads in shared RAN deployments

In the case of the aforementioned large-scale events, if the RAN is shared, public users (e.g., event attendees) can theoretically block radio access to public safety users (e.g., police officers, first aid personnel, firefighters) during a traffic spike. A common situation in this scenario is that during a high peak of network usage, there may be insufficient resources to serve all users with all their demands. Due to the high number of users, another common situation is when uplink interference starts to rise and, to compensate for this, the transmission power of each user starts to increase. This is similar to how people naturally raise their voices in a crowded room to make themselves heard. Under these conditions, users further away from the cell site are at a disadvantage as their transmission experiences higher losses due to the distances involved vis-à-vis users near the cell site.

These situations and others can cause service degradation to both public users and public safety users. However, as mentioned before, the risk to public safety users is greater as a network outage could prevent them from performing their tasks, potentially endangering their own safety as well as that of others.

Thus, considering the venue scenario challenges with high peaks in network traffic and involving RAN shared by public users and public safety users, the main objective of this article is to describe in general terms how to prepare and configure the network to ensure that, during high network load, a minimum level of service can still be provided to the event attendees while ensuring critical communications remain continuously available with a good service level for public safety users.

Mitigating network congestion in shared RAN environments

The first step to ensuring an agreed level of service for both public users and public safety users is to define the expected service levels for each type of user. In the case of public users, this is defined in the contract between the individual users and the commercial operator. In an equivalent way, the public safety operator needs to define, in a contract or a service level agreement (SLA) with a commercial operator, the expected level of service for public safety users under both daily conditions and high load scenarios, and ultimately to determine the specific RAN settings that must be implemented to achieve that goal. As an example, the SLA of the public safety user must ensure the use of operational services such as Mission-Critical Push-to-talk (MCPTT), MC Video or MC Data.

Figure 1 – Preparation for distinct phases of a large-scale event in a venue

Figure 1 – Preparation for distinct phases of a large-scale event in a venue

After a contract or service level agreement is in place, preparation for the event can start. The preparation is divided into five phases. Each phase requires slightly different approaches to prevent, reduce or mitigate service degradation. The five phases are shown in the Figure above and described in the following sections:

  • Before the event: Considerations before the start of the event.
  • Preventive: Considerations to enhance traffic handling before the network is loaded.
  • Reactive (High-load handling): Considerations when the network resources get close to saturation.
  • Reactive (Extreme high-load handling): Considerations when the network is extremely loaded.
  • Continuous: Considerations during the lifetime of the event.

Before the event

Network Design and Tuning

Applying proper end-to-end quality of service parameters and configurations

Challenge: As previously mentioned, public safety users have stricter requirements than traditional public users due to the nature of their work. These requirements are defined in their contracts or service level agreements. During the network design and tuning phase, these requirements must be mapped to appropriate network parameters and configurations to ensure that the service level agreements are met.

It is crucial to parameterize the network beforehand, as making changes during an event is challenging and can negatively impact mission-critical use cases.

Solution: Prior to the event, ensure that proper end-to-end quality of service features and parameterization are in place to prioritize mission-critical traffic during congestion. Additionally, functionality and performance testing must be completed, and any necessary tuning should be applied before the event.

Ericsson offers a Network Design and Optimization service that assists mission critical customers in defining service level agreements and provides guidance on the necessary network parameterization to meet these requirements. The provided Radio Access Network Solution Design includes configurations (dedicated/shared, features, and parameters) to ensure the required mobility, traffic management, and end-to-end user quality of service during both normal load and high-load congestion situations. A post-event deep dive analysis can be conducted to further fine-tune the Radio Access Network Solution Design.

Coverage and capacity planning

Challenge: Venues, whether indoor or outdoor, are often expansive and may feature intricate layouts with multiple floors, walls, and obstacles. Venues can consist of different building materials, structures, and thickness, which affect radio wave propagation and result in various levels of signal attenuation. Ensuring proper coverage across the entire venue, including indoor, outdoor, and below ground areas, can be challenging.

Solutions:

  • Coverage planning: By measuring and modeling signal propagation characteristics, it is possible to identify potential dead zones and thus strategically place antennas and indoor coverage solutions to provide consistent, reliable cellular coverage. During coverage planning, avoiding overlapping cells with the same frequencies is crucial to prevent interference. Careful planning of handover areas is also essential to ensure seamless connectivity.

    Utilizing radio planning tools can streamline coverage planning and reduce the time spent on this. These tools provide a range of functionality and produce different outputs, including coverage maps, capacity reports, interference analysis, performance metrics, scenario simulations, and optimization plans. On-site testing of propagation is crucial to validate the accuracy of simulations generated by planning tools, as real-world conditions may differ from theoretical models.

    Ericsson offers the Ericsson Indoor Planner (EIP) . This is a tool capable of simulating the signal propagation through walls and structures. The tool receives the venue floor plan as an input and creates a plan for the location of small-cell products.
  • Capacity planning: This is a process that involves dimensioning and determining the optimal allocation of radio resources. It involves evaluating and optimizing several factors such as frequency spectrum, capacity requirements, traffic patterns, unicast or multicast, interference levels, and user capacity demand. The user capacity demand should consider the demand of both public users and public safety users.

    During the network Coverage and Capacity planning process, it may become apparent that the current network capacity at the event venue is inadequate to handle the anticipated traffic. In such situations, one solution is to augment the network capacity at the venue by deploying temporary sites, such as cells on wheels, to boost capacity.
End-to-end network resources allocation

Challenge: To transmit and receive information, both public users and public safety users require network resources. Therefore, it is important to ensure that sufficient end-to-end network resources are available to satisfy the service level agreement of all users. These resources can be divided into radio, transport, and core network resources.

Solution:

  • Radio resource allocation: Radio Resource Partitioning (RRP), also called RAN slicing, can be used to control how frequencies are shared in these scenarios, where public users and public safety users occupy the same cell. RRP creates pools or partitions of frequency resources that can be prioritized or reserved for specific types of users. The size of these pools and the user prioritization will depend on the agreed service level agreement. As an example, it is possible to define a pool or partition of frequency resources for which public safety users have priority access. In the case that no public safety users require network resources; these resources will be available to public users. However, under high load conditions, these resources will be prioritized for public safety users. The benefit of RRP is that a minimum set of resources can be reserved for a specific group of users to achieve a minimum level of service.
  • Transport resources allocation: Appropriate dimensioning of the transport network to guarantee that it can cope with the expected traffic demands of both public users and public safety users. Similar to RAN resource allocation, it is possible to reserve transport resources for certain groups of users. This helps to ensure that the service level agreements for public users and public safety users are met.
  • Core resources allocation: Core network slicing allows dedicated resources to be allocated to specific mission critical user groups. For example, it enables a single 5G core (e.g., Public Safety Operator Core) to be divided to serve different user groups within the public safety operator core; each user group can be mapped to a dedicated core slice.
Uplink interference mitigation

Challenge: Reducing interference is key in venue scenarios when considering that the level of uplink interference increases with an increasing number of users and traffic load. During high load scenarios, many users will attempt to transmit back to their cell. These uplink transmissions can cause interference in adjacent cells reusing the same frequency carrier. Another cause of interference is unintended overlap of cells reusing the same frequency carrier. If the cells are not carefully planned, this can lead to interference between neighboring cells. Cell overlaps need to be planned to allow for a proper handover but changes in the venue (e.g., adding or removing bleachers, stages) can modify the cell intended coverage causing interference.

Solution: Proper cell planning and optimization, antenna configuration, and appropriate adjustments to UE transmit power levels can reduce interference between neighboring cells. For instance, reducing the maximum uplink transmission power for public users can create additional capacity for the uplink transmission of public safety users. This approach minimizes the uplink interference caused by public users, thereby improving the performance for public safety users.

Another way to mitigate uplink interference is to use radio features that leverage receiver diversity. Receiver diversity improves uplink coverage and reduces interference by processing multiple channels or paths, thereby enhancing signal reliability and strength.

Network Optimization

Challenge: Because in venue scenarios there are large fluctuations in the number of users, the performance of the network can change drastically. As an example, cell sites covering a venue scenario may normally experience a low amount of traffic and provide a satisfactory service to all users. However, during an event, traffic can increase above the expected levels. Therefore, the network needs to be configured to properly support both scenarios.

Solution: Optimization of cell sites at and near an event venue is needed so that the network is ready for high load peaks. A key goal of the optimization process is to adjust the network parameters to accommodate the anticipated peak load, ensuring that control channels remain safeguarded even during periods of high demand. If control channels are congested, users (including event attendees and public safety) cannot access the network even if there are free data channel resources. Therefore, it is important to protect the control channels to guarantee that all the admitted end users are served when needing to access and use the network.

Preventive (Low load) handling considerations

Traffic steering (Offload, Load-balancing)

Challenge: It is possible to have multiple frequency carriers in a single cell site. As an example, the operator may have a mix of carriers in low and mid-band frequencies in a single site. Because bands have distinct characteristics (e.g., propagation, available bandwidth), the challenge is for the network to steer users to the most appropriate frequency carrier considering the type of user, their service demands, and the carrier load.

Solution: Handovers are initiated based on measurements of signal power (e.g., Reference Signal Received Power), interference levels (e.g., Reference Signal Received Quality), or the load on neighboring cells. During handovers, traffic steering can be employed to direct users to the most suitable cell, such as the cell with the lowest load or the best propagation characteristics. This steering can be managed independently for public users and public safety users.

Traffic steering can also be employed for "band clearing," which involves freeing up specific frequency bands for the exclusive use of public safety users, before an event. This ensures that, in addition to the bands shared with public users, public safety users have a dedicated band during the event.

Reactive (High load) handling considerations

User access prioritization

Challenge: As mentioned before, public safety users have different requirements than public users. If a public safety user fails to connect to the network, this can imply a safety risk. Thus, even under high loads, public safety users must be able to access the network.

Solutions:

  • Access Control: Each user has been assigned an Access Class or category stored in the SIM card. The cell transmits, as part of its broadcast information, the access classes or categories allowed in the cell. A device cannot send any connection request to the cell if its access class or category is not announced by the cell. Because public safety users have a different access class or category than public users, the network can restrict public users’ access to the cell in cases of high load, while still allowing access to public safety users. The use of access control (e.g., Unified Access Control, Access Class Barring Control) helps to ensure that public safety users have access to the cell and that existing connected users can have a minimum quality of service.

    Because Access Control can limit public users access to the cell (in order to protect public safety users from network congestion), it is crucial to tune this feature carefully to avoid unnecessarily blocking public users.
  • RRC Establishment Cause: In a connection request sent by a user, using the Radio Resource Control (RRC) connection procedure, a RRC connection cause is specified. The cause indicates if it is a request from public safety users (i.e., identified by its access class or category) or from any public user making an emergency call. This allows the network to give priority to public safety users or even to public users attempting to make emergency calls.

Admission control and preemption

Challenge: Assuming that all cell resources have been consumed due to high load, but a user has been admitted in the cell because it is a public safety user, or public user attempting to make an emergency call, the network needs to allocate resources to this user. To do that, the network should be able to pre-empt public users with lower priority to obtain resources for public safety users or any user attempting to make an emergency call.

Solution: Admission Control (part of RAN quality of service features) can be used to prioritize users by managing network resources and ensuring that high-priority traffic (i.e., public safety users and public users attempting to make emergency calls) obtain the necessary bandwidth and performance. In low load conditions, all users can access the shared/assigned resources. However, during high load situations, if a high priority user is admitted to the network and there are not enough resources, the network can pre-empt lower priority public users to recover sufficient resources for higher priority users.

Traffic steering (Offload, Load-balancing)

Same as above.

Reactive (Extreme high load) handling considerations

UE power control

Challenge: As mentioned in a previous section, during a high load scenario, users will tend to raise their transmission power to the maximum to overcome uplink interference from other user transmissions. However, users far away from the cell will be disadvantaged compared to those near the cell site. As both are transmitting at the same maximum power, the UE closer to the cell can interfere with the transmission of the user located farther away. As an example, if a public safety user’s UE is located far from the cell site, when it attempts to communicate with the cell site, its transmission may not be detected as it will be drowned by the more powerful transmissions of other users closer to the cell site.

Solution: In a high load scenario, it is possible to reduce the transmission power of public users. This will give an advantage to public safety users. Then, if the UE of a public safety user is located far away from the cell site, the higher transmission power of the UE of the public safety users will compensate for the longer distance compared to the public users closer to the cell site. In 5G, by using more advanced power control features, it is possible to differentiate between groups of users, and set different uplink transmission powers to these distinct user groups.

Traffic steering (Offload, Load-balancing)

Same as above.

Admission control and preemption

Same as above.

User access prioritization

Same as above.

Continuous aspects during the lifetime of event

Performance monitoring

Challenge: Due to fluctuations in the numbers of users and their demands on the network, the performance of the network may change suddenly. Therefore, it is necessary to constantly monitor the network performance to detect congestion.

Solution: Monitoring of real time performance management counters, key performance indicators (KPIs) and alarms. For public safety users, voice quality is critically important. Therefore, key performance indicators related to voice quality, such as MCPTT KPI 3 (mouth-to-ear latency), need to be closely monitored.

KPIs are divided into Resource-KPIs (R-KPIs) and Service-KPIs (S-KPIs):

  • Resource KPI (R-KPI): The Resource KPIs measure the performance of network domains and network elements. They are good indicators of system capability and may also be used for troubleshooting and performance degradation identification and localization.
  • Service-KPI (S-KPI): User-perceived service performance that is monitored on a per user basis.

Network Optimization

Same as above.

Conclusion

Large scale event venues are characterized for having large variations in the number of users. This is a challenge for the radio network design that needs to maintain connectivity and services to different classes or types of users, even during high peaks of network usage.

Public safety operators can and often do share the RAN with commercial operators to lower overall costs and accelerate deployment timescales.

A RAN shared between a public safety operator and a commercial operator can provide good coverage to public safety and public users in a venue scenario during an event. However, it is important that there is a contract or service level agreement between the public safety operator and the commercial operator describing the minimum service level required for public safety users during normal and high load conditions.

Furthermore, a suite of mechanisms can be applied in distinct phases of the event to ensure that both public users and public safety users are served according to their needs and agreed service level requirements. The table below summarizes the mechanisms required to effectively handle the high-load events described in this paper.

  Key considerations Event timeline
Mechanism Interference Prioritization Resource Allocation Before Event Preventive
(Low-Load)
Reactive
(High-Load)
Reactive
(Extreme high load)
Continuous
End-to End QoS Design and Config. x x x x        
Coverage and Capacity Planning x   x x        
E2E Resource Allocation     x x        
Uplink Interference Mitigation x     x        
Network Optimization x     x       x
Traffic Steering
(Offload, Load-balancing)
  x x   x x x  
User Access Prioritization   x       x  
Admission Control and Preemption   x       x x  
UE Power Control x           x  
Performance Monitoring x x x         x