At Mobile World Congress Americas, Sprint and Ericsson announced the results of the first U.S. 2.5 GHz Massive MIMO (multiple input, multiple output) field tests conducted in Seattle, Washington and Plano, Texas using Sprint’s spectrum and Ericsson’s 64T64R (64 transmit, 64 receive) radios. The two companies are preparing for commercial deployment next year, with Massive MIMO radios capable of increasing Sprint’s network capacity up to ten times.
Dr. John Saw, Sprint CTO, said, “Massive MIMO is a tremendous competitive advantage for Sprint, enabling us to maximize our deep 2.5 GHz spectrum holdings.”
Testing of Massive MIMO on the Sprint LTE Plus network in downtown Seattle showed a capacity increase of approximately four times compared to an 8T8R antenna. To showcase this capacity, Sprint convened 100 people with Samsung Galaxy S7 phones and ran simultaneous file downloads on a timed-test on all networks. The testing showed a 100 percent success rate on the Massive MIMO-powered Sprint network.
In Plano, Texas, Sprint and Ericsson also recently tested Ericsson’s 64T64R Massive MIMO radios reaching peak speeds of more than 300 Mbps using a single 20 MHz channel of 2.5 GHz spectrum.
For both field trials, Ericsson provided the radio network infrastructure and backhaul equipment. Sprint and Ericsson together developed the test cases and requirements, which included a variety of performance scenarios involving multi-user and non-stationary testing. The Radio Network infrastructure included Ericsson’s next-generation 5G-ready AIR6468 radio, and the backhaul equipment utilized the MINI-LINK 6352 R2 microwave radios which can provide up to 10 Gbps of backhaul, future proofing the network for 5G.
July 18, 2017 —
While the Washington Nationals easily handled the Seattle Mariners during a recent game, the real contest was between performances of the cellular networks and the Wi-Fi system as they battled to provide the best online experience.
During a recent major league baseball game at Washington Nationals Park, independent network benchmarking firm Global Wireless Solutions (GWS) conducted customer experience mobile network testing. The tests revealed that, both before and during the game, the Wi-Fi system excelled in the area of speed but fell behind cellular when it came to reliability.
Paul Carter, CEO GWS, said, “People want a reliable network with reasonable speed that works when you make a call, post a selfie, or load a video. If you’re at a major sporting event, you want to quickly take care of your online activities and watch the game not your phone.”
Before the game, the carriers’ data speeds for a 4 mb file upload (the size of a Snapchat video) ranged from 1 to 4 Mbps, according to Global Wireless Solutions, while the Wi-Fi network averaged speeds of 8 Mbps.
AT&T, Sprint and T-Mobile also provided consistent speeds throughout the evening, while Verizon’s speeds dropped sharply during the game, according to GWS’ testing. For example, Verizon’s download speeds for watching a short video clip fell from 2.8 Mbps before the game to 1.6 Mbps during the game, before rising to 4.2 Mbps after the game
Wi-Fi Speed Crushes Cellular
During the game when the Park was the busiest, the Wi-Fi network was capable of delivering an average of roughly 32 Mbps, while the fastest cellular network, AT&T, averaged 25 Mbps.
“The $300-million program to bring Wi-Fi to every major league baseball park in the United States has brought in-seat connectivity into the 21st century,” the firm wrote. “When measuring potential capacity download throughputs, the Nationals Park Wi-Fi network was overall higher than those measured on cellular networks.
Cellular Owns Reliability
However, while the Park’s Wi-Fi was the quickest, it was not the most reliable. All the carriers were more reliable in completing data tasks, nearly 100 percent, while Wi-Fi was several percentage points behind.
“A consistent Internet experience is highly valued. Steady with reasonable speed is a better experience than a network which is fast, then becomes too slow to undertake some common tasks, then suddenly speeds up again,” Carter, said. “If you want to share a photo or send a video, you want your network to support that dependably. For some baseball fans, the WiFi network in the Park can provide a better experience than their own LTE connection.”
For voice calls AT&T, Sprint and Verizon all had 100 percent reliability with AT&T and Verizon using VoLTE the entire time. T-Mobile, also using VoLTE, wasn’t far behind, however, it did experience 1 in 12 calls failing before the game started.
June 21, 2017 —
Using Massive MIMO Samsung radios, equipped with vertical and horizontal beam-forming technology, Sprint achieved peak speeds of 330 Mbps per channel using a 20-megahertz channel at 2.5 GHz during field trials in South Korea. Capacity per channel increased about four times, cell edge performance increased three times and overall coverage area improved as compared to current radios.
Günther Ottendorfer, chief operating officer – technology, at Sprint, said. “Massive MIMO is a tremendous differentiator for Sprint because it is easily deployed on 2.5 GHz spectrum due to the small form factor of the radios needed for a high frequency band. In lower frequency bands, wavelengths are much longer and therefore the radios require much larger, impractical form factors. This makes Massive MIMO an important tool for unleashing our deep 2.5 GHz spectrum holdings.”
The Massive MIMO radios use 128 antenna elements (64 transmit, 64 receive) compared with 16 elements 8T8R antennas 8T8R (8 transmit, 8 receive) radios currently deployed by Sprint across its U.S. network. The purpose of the test was to compare the performance of Massive MIMO radios with 8T8R radios. The testing included multi-user and mobile user cases. Samsung provided the Massive MIMO network infrastructure, as well as the test network design, operation, data collection and processing. Both companies will use the results in preparation for commercial deployment of Massive MIMO in the U.S. and other markets.
In cities across the U.S., Sprint plans to deploy Massive MIMO radios with 128 antenna elements using its 2.5 GHz spectrum. In March, Sprint deployed Gigabit Class LTE on a live commercial network in New Orleans. There Sprint used three-channel carrier aggregation and 60 megahertz of 2.5 GHz spectrum, in combination with 4X4 MIMO and 256-QAM higher order modulation, to achieve Category 16 LTE download data speeds on a TDD network. With Massive MIMO radios using 64T64R, Sprint has the ability to push capacity beyond 1 Gbps to reach 3-6 Gbps per sector.
May 11, 2017 —
Working with Qualcomm Technologies and SoftBank, Sprint is developing 5G technologies in the 2.5 GHz band, including the 3GPP New Radio (NR) standard.
“Today 2.5 GHz TDD-LTE is one of the largest global wireless ecosystems used by some of the most influential operators in the world such as SoftBank and all of China’s operators, including China Mobile,” John Saw, Sprint CTO, wrote in a blog today.
The agreement includes developing a 3GPP 5G New Radio (NR) on the 2.5 GHz spectrum, which Sprint expects to deploy in late 2019.
“As one of the first proponents of 2.5 GHz and TDD-LTE for 4G, we understand the value of building a strong global ecosystem early on. This is why we are working with Qualcomm and SoftBank to develop the 3GPP 5G NR capabilities for 2.5 GHz,” Saw wrote.
Sprint has more than 160 MHz of 2.5 GHz spectrum in the top 100 U.S. markets, more mobile-ready 5G spectrum than any other U.S. carrier.
“Today we’re using this advantage to densify our network with more cell sites and antennas to build a strong foundation for 5G,” Saw wrote. “We are ensuring that Sprint’s deep 2.5 GHz spectrum is an early first-mover in the 5G ecosystem. Not all spectrum bands have this kind of global support and economy of scale.”
Last December, Sprint announced that it had developed technology that will increase 2.5 GHz network coverage by up to 30 percent to nearly match 1.9 GHz spectrum performance, while penetrating buildings at a rate of be 90 percent of what is achieved at 1.9 GHz spectrum.
The carrier demonstrated the advanced technology, known as High Performance User Equipment (HPUE), which a new power class – Power Class 2 – for end-user devices such as smartphones. In development for two years, HPUE was designed to improve the performance of TDD-LTE Band 41 networks, John Saw, wrote in a blog.
“In 2015, we began working on a solution to improve our 2.5 GHz coverage by increasing the uplink coverage of Band 41 devices,” Saw wrote.
Sprint said that its network is ready for the initial rollout of HPUE in its 250 LTE Plus markets. Samsung, one of Sprint’s most important ecosystem partners, is expected to support HPUE in devices slated for commercial launch sometime this year.
Postpaid net additions grew by 347,000, a 54-percent increase from 62,000 added in FY2Q15. These adds bring the total postpaid subscriber base to 25.7 million, up 1 percent on a year-to-year (YtY) basis. At the same time, customer retention programs are working; postpaid churn came in at 1.37 percent, down from 1.49 percent in FY2Q15. Postpaid average revenue per user (ARPU) keeps sliding, however, to $50.54, down 2 percent from $53.99 in FY2Q15.
Price competition with the other Tier 1 carriers has taken a toll on the top line, however. The net effect is that postpaid service revenues at $4.7 billion were flat to down 1 percent from $4.9 billion a year ago.
In turn, Sprint’s ability to fund its network expansion plans is hampered as it curtails capital expenditures (capex) to conserve cash.
Certainly, the company is reiterating a strategic program to optimize and densify its network. This program involves:
It is this latter point that raises some serious concerns as to how Sprint will achieve its network expansion goals and drive new revenues.
Sprint’s quarterly wireless capital spending has been in a nosedive for the past six quarters, declining at a compounded rate of 26 percent per quarter. From a peak of $1.6 billion in FY1Q15 (calendar 2Q15), Sprint’s wireless spending dropped steadily throughout the year to $577 million in FY4Q15. Wireless capex for the year totaled $4.1 billion, down 16 percent from$4.9 billion in the prior year.
The capital spending slide continued in the current fiscal year. FY1Q16 wireless capex came in at $376 million and dropped further to $358 million in FY2Q16.
It is important to note that capital efficiency (that is, the ratio of capex-to-service revenues) for these periods hit only 6 percent. That is barely in maintenance mode. When networks are expanding, carrier capital efficiency historically is at 15 percent or more. Even with capital cost saving measures, we would expect Sprint’s wireless capital efficiency ratio to be in low double digits, just to be able to meet its plan.
The big question is how much the company will invest through the balance of its current fiscal year.
In its October 25 earnings call, Sprint offered full-year FY2016 guidance for cash capex of “less than $3 billion, excluding devices leased through indirect channels.“
If we drop out the portions for wireline network and corporate capital expenditures, we estimate that Sprint has budgeted around $2.7 billion in wireless capex for FY2016.
Through mid-year, Sprint spent $734 million or just 27 percent of the budgeted total.
If it sticks to its plan, this means that Sprint must spend the nearly $2 billion balance on its wireless network upgrade and expansion over the next two quarters.
Can Sprint do it? How will it do it?
John Celentano is a technology marketing consultant and a wireless infrastructure expert.