Approaching Hi-Cap Venue DAS Optimization
Introduction: Over the years wireless network traffic has exploded. To meet this demand operators have focused efforts toward building (Distributed Antenna System) DAS Systems at high capacity venues. For a single event like Super Bowl, a stadium may have over a terabyte of data passing through the serving cells. Building and optimization of a DAS is a challenging process which requires equipment validation, DAS configuration adjustments, surrounding network interference mitigation, RAN network parameter optimization, and neighbour and Layer management optimization to support heavily loaded conditions.
Background: Events have an extreme effect on the increase in network traffic at high capacity venues due to the nature of use. Games, concerts, and such lend to today's social media update mentality driving this explosive data utilization. This increase in demand for network traffic had led to the expansion of networks closer to the users and build out of iDAS, COW’s, and oDAS systems to augment the existing Macro Coverage surrounding an event. For network operators it’s particularly important for their subscribers to have a positive network experience. During a peak hour at a stadium which with 100,000 subscribers the traffic requirements exceed the general requirements of a city with 500,000 customers. What makes it a further challenge is most of the users utilize the devices at the same moments resulting in an extremely high demand for network service in a concentrated time frame. Optimization of these high capacity DAS systems are important to ensure the RF quality at the venue is maintained as well as the system is able to handle the heavy traffic load during event hours.
Mobile operators face continuous challenge to drive improvements to the services they provide. This extremely rapid growth in network places greater demands on coverage and capacity. With Mobile traffic exceeding desktop traffic at exponential rate operators are rushing to fill up this demand by build out of dedicated systems like DAS to address coverage and capacity at the same time keeping the infrastructure costs low. This case study explores the challenges presented to the network operator in the optimization of the DAS systems at high capacity venues.
Optimization Process: Building of a DAS system is only part of the capacity requirements. Spending millions on the system doesn't equate to anything if it not performing up to the designed capacity requirement. This is where venue optimization is critical. Optimizing a venue is not an adhoc process, rather a systematic approach to ensure all aspects of the DAS and venue performance are operating in harmony.
Figure 1: General HiCap Venue Optimization Process
Before the venue and DAS is fully ready to be optimized, it is critical to ensure that there are no hardware issues and the site is commissioned as per designed. Before validation testing is conducted, it is important to run through all alarms on DAS and the eNodeB level at the site. UL RSSI reports for all sectors are run and made sure they are within the specified range for the DAS. Validation testing is to be conducted by walking under every antenna of the venue for all technologies and carriers.
Here are the important things to check for during validation:
- Ensure all Antennas are transmitting at proper power level
- Make sure all sectors are transmitting correct PCI/PSC/PN/ARFCNs
- Make Test calls and data sessions in all sectors to ensure all sectors are taking traffic
Validation testing results should be documented for future reference and analysis
Figure 2: Example Antenna Validation Table
After the DAS has been validated that there are no hardware issues, the venue is ready for optimization. Before commencing the site walk for optimization, the following items should be taken care of.
- Make sure all neighbours between DAS sectors are defined
- Ensure 1st Tier neighbours with Macro sites are defined
- The site has the correct software load for a high capacity venue and conduct an audit of parameters to verify correct high capacity parameters are defined
- Study the design files and identify areas of potential issues (i.e. coverage holes, high handover zones, and overshooting sectors)
Walk testing is conducted for all technologies for call, data and scan (all carriers) for every 4th to 5th row for a stadium to get detailed RF Coverage maps. Walk testing data is utilized for the following objectives:
- RF propagation plots for coverage and quality
- Ensure that the design thresholds are met
- Identify strong macro penetration and create strategy to reduce macro impact
- Coverage footprint for all sectors and identify overshooting sectors
- Determine a baseline for walk test KPIs.
- Voice/Data Accessibility.
- Voice/Data Retainability
- Data Throughput (DL/UL)
Zone Optimization: One of the greatest challenges in improving of SINR for a high capacity venue is containment of coverage from one zones nodes overshooting into other zones. Poor zone containment increases resource utilization and adds system self interference impacting the SINR seen by the customers devices. This issue is further pronounced in closed arenas and stadiums where many sectors shoot into the field or "bowl". Figure 3 demonstrates this scenario where numerous sectors shoot into the bowl area and over propagate into other sectors zones impacting the SINR.
Figure 3: Overshooting Sectors within DAS Cells
The overshooting sectors impact the SINR which in turn leads to users operating at higher power leading to increase in system noise and leading to further degradation of network performance. RF shaping for each zone is utilized based on the following general strategies:
- Adjust Gain/Attenuation at sectors level to ensure every sector is only dominant in its area of intended coverage.
- Tilts need to be employed and directional antennas need to be adjusted to reduce the coverage footprint of overshooting sectors.
Walk Test Optimization - Loaded Conditions: Although the optimization of the DAS is usually conducted in non-traffic condition for RF Shaping and macro reduction, to simulate the event conditions the DAS is put through a stress test by placing an artificial load on the network through OEM OSS settings. An additional walk test should be conducted in loaded conditions. Loading of the DAS shouldn’t exceed 50 % of System load M-Pole. If the MPole is increased by 50 %, a 3 dB Noise rise can be expected.
Figure 4: OSS Simulated Loading
Additional walk testing should be conducted to ensure that the primary KPIs are within accepted levels as defined by the operator for the venue.
Macro Reduction: The most important aspect in designing of an Indoor DAS (iDAS) is ensuring the nodes coverage is greater than the surrounding macro cells by at least 7dB. This requirement is essential in making sure the SINR requirements are met and also to ensure the traffic is offloaded to the DAS network. To ensure the macro layer coverage does not impact the DAS performance, macro reduction exercise should be conducted to ensure good isolation of DAS over macro without impacting the outdoor coverage. For details on Macro Reduction using tools such as OptPCS-ICE and the cost savings associated with it refer to
Figure 5: Overshooting Macro Cells within DAS Cells
Parameter Optimization: The RAN network parameters being set correctly are vital to balancing load, minimizing interference, and directing customers to the best layers. As important as it is to have parameters set completely, it must also be understood that the parameters can not overcome poor design, integration, and cell coverage optimization. Parameter optimization should take into account the following:
- Access parameters (PRACH)
- SHO Strategy for 3G to reduce the SFO factor and lower resource utilization
- Capacity offloading Strategy to ensure traffic is balanced between all network layers
- Inter frequency and Inter System HO Strategy to reduce time in compressed mode
- Idle mode parameters to direct subscribers in a way to distribute traffic across bands and neighbouring cells
- High capacity parameters for maximum traffic handling by the network
Live Event Optimization: During the events duration, KPIs are collected from the OSS at the venue. Carrier and sector level statistics are measured to ensure that the performance is within acceptable levels. If any sector or site is overloaded, the traffic parameters are adjusted to ensure that the traffic is balanced between different carriers and there is minimal congestion on any site. Some of the important KPIs to ensure are acceptable as defined by the operator.
- Accessibility – CS and PS
- Retainability – CS and PS
- Throughput – UL and DL
- HO Success Rate
- IRAT% and SRVCC%
Applied Application - Super Bowl 2015: University of Phoenix Stadium outside Phoenix, AZ hosted 2015 NFL Super Bowl. During the event the number of users exceed 100,000. To serve the number of users, a network was designed with the following cell for LTE and 3G:
- iDAS Sectors – 38 sectors within the venue
- Temp Sites – 4 Cell on Wheels (COWS) surrounding
- oDAS System – 17 Sectors with 16 Nodes Surrounding Venue
- Macros – 4 Surrounding Macro Cell Sites
For the optimization of a complex venue like this requires a comprehensive strategy and complete execution of the outlined activities for HiCap DAS Optimization. The Super Bowl activities were complete with these activities.
- DAS node validation is conducted to ensure all antennas were transmitting with proper power levels and any hardware issues are identified and fixed. Coverage thresholds were verified at venue level and sector level to ensure the DAS built is as per designed specifications. These were compared against the design predictions.
- Macro reduction is conducted using ICE to reduce impact of surrounding COW’s and Macro sites. Macro reduction exercise is conducted in multiple phases to capture all macro interactions from a three tier boundary, changes made to the macro sites are made by ensuring there is no loss of on street coverage.
- All high cap parameters were aligned to handle maximum traffic. Neighbours lists and priorities were set utilizing OEM user trace recordings during previous games to capture detected/missing neighbours
- RF zone optimization was conducted having all zones contained through adjusting tilts and attenuation on the DAS antenna nodes to ensure sectors do not over propagate beyond their designed coverage area into other sectors impacting venue SINR.
- Walk bench marking and optimization with data collection performed with and without loading to determine the impact on the primary network performance KPIs. Results were compared against primary competitors.
Figure 6: Super Bowl Venue Zone Validation
Figure 7: Detailed Zone Optimization
Results: During the day of the Super Bowl, the operators data traffic exceeded 1.2 Terabytes all the while performance KPIs for the venue remained solid providing a great overall customer experience.
|Site Type||CS Access||CS Retain||PS Access||PS Retain||DL Tput(kbps)||DL Vol (GB)|
Table 1: Performance During Time of Super Bowl Event
Takeaways: Venue optimization comprising of DAS, macro, and other transmitting nodes as complete system is key to quality venue performance for your customers. In order to meet traffic demands of high capacity events, extra nodes of a variety of types are required. Planning each separately with a complete plan of optimization post launch will allow for improved network performance and overall cost savings. Without a complete plan and just utilizing a DAS "as is" often results in a loss of money both through return on investment for capacity as well as negative net promoter scores from subscribers.
RFAssurance is a Telecom Technology Services, Inc department specializing in the support of wireless networks. RFAssurance provides support and consulting for RF RAN and Core Network tools, processes, and results to improve network design, optimization, and general performance. Our managers, engineers and software developers are subject matter experts with various design & optimization tools, database structure, web implementation, and their practical applications. For how we can help support your network, contact us via email at firstname.lastname@example.org