pCell® vRAN is a software-based Virtual Radio Access Network that uses Artemis pCell technology to deliver more than 10 times the aggregate data rate per megahertz of a cellular RAN/vRAN uniformly throughout the coverage area, while remaining compatible with unmodified LTE and 5G devices.

pCell vRAN runs on AMD- or Intel-based COTS (Commercial Off-The-Shelf) servers, without hardware acceleration, connected toPoE (Power Over Ethernet) based Artemis pWave® RRUs (Remote Radio Units) placed throughout the coverage area wherever it is convenient, using Ethernet cabling and switches.

While a conventional cellular RAN/vRAN carefully divides the coverage area into cells, avoiding interference, pCell vRAN exploits constructive interference to synthesize a personal cell (“pCell”) for each user device concurrently, allowing all pCell users to use the full capacity of the spectrum at once. As users move throughout the coverage area, each user’s pCell remains locked their user device, providing continuous, near-peak downlink and uplink data rate to all users at once.

Unlike a cellular RAN/vRAN, pCell vRAN is cell-free, with no handovers, no cell edge conditions, and no intercell interference. And unlike cellular, regardless of how close users are packed together—even if phones are stacked—each user device gets an independent pCell.

pCell vRAN includes a built-in EPC/5GC and optionally supports an external EPC/5GC through standard interfaces.

pCell vRAN is a managed, subscription-based service. Artemis or an authorized system integrator will work with you to dimension and install pCell vRAN servers and pWave RRUs to meet your needs, using existing cabling and racks where possible.

Artemis will work with you to provision eSIMs and/or SIMs with your chosen branding and optional roaming on national networks.

For permanent installations, Artemis will charge an installation fee and an ongoing subscription.d. For temporary installations, such as music festivals, Artemis will charge an installation and usage fee for the duration of the event.

If you opt for Collective Roaming (mobile service beyond the pCell vRAN coverage area using national mobile operators), Artemis will charge you based on overall megabytes of roaming usage.

Artemis will retain ownership of the pCell vRAN servers and pWave RRUs, and recover them once the subscription (or the temporary event) comes to an end.

Absolutely. Choose whatever name you'd like to appear as the network operator for your pCell Private LTE/5G network. In fact, you can have multiple different network operator names appear on phones, for example, for different kinds of users that perhaps you configure with different classes of service.

As another example, a sponsor might pay a venue for its name to appear as the network operator, and perhaps even pay for Collective Roaming service for, say, VIP season ticket holders.

You can set up your pCell vRAN offering anyway you want.

Collective Roaming is what Artemis calls an optional service where your Private Network users can roam beyond the free CBRS spectrum in the pCell vRAN coverage area and continue to have service through a national mobile operator for a per megabyte roaming fee. We call it “Collective” Roaming because, rather than setting up a separate national operator account for each of your users—many of whom may never or rarely roam—you can have one roaming account for all of your users collectively, and only be billed for roaming that is used.

For example, consider a college campus with Multi-Gigabit pCell vRAN coverage throughout the grounds, buildings and sports facilities using free CBRS spectrum. Students and faculty, who collectively spend 90% of their time on campus, incur no mobile operator cost while on campus. With Collective Roaming, the college can offer a service to students and faculty where anyone who roams off campus will continue to have mobile service through a national operator, and since that is only 10% of the overall mobile usage, the college would be billed a fraction of what it would have cost if each student or faculty had individual national operator accounts. Whether the user is on-campus or roaming, their phone will still show the college name as the network operator.

It's up to you, the pCell vRAN customer, whether a college, enterprise, venue, or healthcare facility, how and whether you charge users for your pCell vRAN network, either when using free CBRS spectrum or (if activated) when using Collective Roaming.

Yes, less than 1/10th the power, despite delivering much higher and more uniform data rates, resulting in a much lower carbon footprint.

As example, for pCell vRAN to deliver 1 Gbps in 20 MHz throughout the 140,000 sq. ft. (13,000 m2) 20,000 seat SAP Center arena:

There are 3 servers, each drawing less than 750W (3 * 750 = 2.25kW), and as a rule of thumb, we double the server power consumption to account for the power to run the room air conditioner to keep them cool. This brings the server power consumption up to 4.5kW, and let’s round up to 5kW to allow for Ethernet switch power and the 28 pWave RRUs each drawing less than 7W (28 * 7 = 196W).

In contrast, the SAP Center cellular DAS is a large room full of many racks of equipment, and there are also 32 large active DAS RRUs in the arena. According to this reference the DAS Head-end and BTS Hotel consume between 18kW and 120kW, and the DAS RRUs consume between 400W to 1.8kW, so for 32 RRUs, that is between 12.8kW and 57.6kW. Summing the DAS Head End/BTS Hotel and BTU power and doubling the totals to account for air conditioning, in total, the cellular DAS RAN consumes between 62kW and 355kW.

In conclusion, the SAP Center pCell vRAN consumes about 5kW. The cellular DAS RAN consumes between 62kW to 355kW. Despite consuming a fraction of the power, pCell vRAN delivers a much higher, more uniform data rate than the cellular DAS RAN.

A Private LTE/5G Network is a mobile network that is for your own use rather than as part of a public mobile network. Private LTE/5G provides mobility throughout the coverage area withSIM/eSIM-based security and full IT control of users, data, networks and policy.

Artemis pCell Private LTE/5G Networks uniquely offer multi-gigabit, uniform performance throughout the coverage area as a turn-key, end-to-end, completely software-based vRAN solution. While conventional Private LTE/5G Networks provide <100 Mbps in 20 MHz of spectrum, pCell Private LTE/5G networks provide >1 Gbps in 20 MHz and>7.5 Gbps in all 150 MHz of CBRS spectrum where available.

In the United States, you can use free CBRS spectrum for Private Networks, typically allocated in 20 MHz blocks. Other countries have also made available spectrum for Private Networks in one form or another, sometimes for a fee. It is also often possible to lease licensed spectrum for aPrivate Network. 

In general, Artemis charges an initial fee to install servers and pWave radios, and then a monthly fee based on the aggregate data capacity of the pCell vRAN. There are additional fees for eSIMs/SIMS, and if you choose roaming or other services.

But each customer and each deployment is different and there are often specific requirements. Once we understand your requirements, we can give you a quote.

No. pCell vRAN is a managed, subscription-based service. It is not available for sale.

The pCell vRAN servers and pWave RRUs remain the property of Artemis and are returned when the subscription ends. Artemis may also swap in newer pCell vRAN servers and/or pWaves over time.

The process begins with the customer contacting Artemis through the Contact Us page on this website and describing their needs. Artemis will then reach out to you to discuss your requirements in detail and get an understanding of your dimensioning requirements and special needs.

Artemis or an authorized system integrator will then meet with you are your site and then typically will set up a temporary pCell vRAN installation and distribute phones throughout the coverage area to demonstrate the pCell vRAN performance compared to existing cellular RAN/vRAN performance.

Artemis or the system integrator will then install the pCell system.

Artemis will then work with you to provision eSIM or SIM cards for your user, and to see whether you’d like eSIM installed via an app, a QR code or both.

After that, Artemis will continue to manage the pCell vRAN in accordance with your needs.

Artemis will respond as soon as it can. We are transitioning from beta to commercial operations and we appreciate your patience.

Depending upon the size of the location, the difficulty of installation and the dimensioning required by the customer, a pCell vRAN can be temporarily setup for testing within a week, and a permanent installation can be installed in weeks.

In the case of temporary events, such as music festivals, a temporary pCell vRAN can be set up in days that can provide service to not only the concert attendees, but to performers, crew, ticket takers, vendors, camera operators and other staff.

pCell vRAN requires much less time than a cellular RAN/vRAN to install because there is no cell planning and pWave RRUs can be installed wherever it is convenient to attach them and provide them with an Ethernet connection. Since pCell vRAN uses conventional asynchronous Ethernet, no specialized cabling or switches are required, whereas IEEE 1588, SyncE, CPRI, or CPRI over Ethernet are typically required for cellular RANs/vRANs.

From invention, through development, to manufacturing, and to final assembly, the pCell vRAN and pWave RRUs have been created entirely in the United States by an incredibly dedicated and versatile team. Almost all key components are from United States companies.

No. pCell vRAN includes a built-in EPC/5GC and optionally supports an external EPC/5GC through standard interfaces.

pWave RRUs support all mobile LTE/5G bands from 600 MHz to 6 GHz as configured by the pCell vRAN software. pWave RRUs will drive antennas supporting any band within this range.

In TDD bands, pCell increases both downlink and uplink capacity by more than 10 times. In FDD bands. pCell increases only uplink capacity by more than 10 times.

Because wireless links (transmissions) in cellular LTE/5G, and Wi-Fi networks can be easily intercepted from any location within range of a cell or base station, cryptography is used to protect the transmitted data. If a bad actor gets access to crypto keys (e.g., via a hacking exploit or a rogue employee) they will be able to intercept all transmitted data.

In contrast, if a pCell vRAN downlink transmission is intercepted from any location other than the exact 6-dimensional (x, y, z, yaw, pitch, roll) location of the intended user, only noise will be received. Similarly, if an Ethernet cable transmission to a pWave RRU is intercepted, only noise will be received. We call this pCell Physical Layer Security (“PPLS”). “Physical layer” refers to the actual medium carrying a signal, be it the copper or fiber in an Ethernet cable, or radio waves through the air.

PPLS inherently securely transmits downlink data to user devices. A random private key can be securely transmitted to user devices to be used for private encryption of uplink data.

PPLS is in addition to standard LTE/5G encryption.

While PPLS adds a significant additional obstacle for bad actors seeking to intercept LTE/5G transmissions, it’s an entirely new approach to wireless security and, like any new security technology, Artemis cannot yet guarantee it can’t be circumvented. (If you are a white hat hacker and figure out a way, please let us know.) As always, continue to protect the cryptography keys used by your LTE/5G network and, for now, think of PPLS as another measure of protection.

No. While pCell vRAN servers are COTS (Commercial Off-The-Shelf) servers, they are specially configured for the ultra-real-time requirements of the pCell vRAN software.

Not using standard Virtual Machines (VMs). Existing VMs introduce execution delays and cache, memory, bus, and network overhead that, while small for most applications, are much too large for the required ultra-real-time performance that pCell vRAN software must maintain constantly to generate real-time radio waveforms entirely in software.

That said, pCell vRAN software is running in a new kind of ultra-real-time, multi-thread/-core/-CPU/-server VM developed by Artemis with far less overhead, and microsecond-resolution deterministic execution and resource utilization.

In the future, Artemis will provide APIs for this VM for ultra-low-latency Multi-Access Edge Computing (MEC) applications, such as Extended Reality/Augmented Reality. This will enable MEC applications to run on both the same servers as the pCell vRAN and on other local edge servers with sub-20 µsec latency.

pCell vRAN servers are AMD- or Intel-based COTS (Commercial Off-The-Shelf) servers. The power and environmental requirements depend on the particular COTS servers being used in a given installation.

As an example, the three pCell vRAN servers currently installed at SAP Center that deliver over 1 Gbps in 20 MHz are AMD Epyc-based 1U COTS servers, each with a conventional Ethernet NIC. Each server uses less than 750 Watts. They are installed in SAP Center’s air-conditioned sound equipment room.

pCell vRAN software is very flexible and scalable, and can run on any COTS servers, e.g., ones that are aviation certified for pCell vRAN-based mobile service on jets, or for small-scale deployments on small AMD- or Intel-computers like NUCs.

Each pWave RRU is 7.4x4.7x1.5” (18.8x12x3.9cm) and weighs 1.9 lbs. (0.85 kg.). Each pWave is powered through PoE (Power Over Ethernet), drawing less than 7W.

Yes. pCell vRANs support all carriers, all mobile bands, LTE and 5G protocols, and through simple software configuration can host carriers in either their own spectrum or in unlicensed spectrum, such as the CBRS band in the United States.

When carriers activate pCell vRAN as a neutral host, their subscribers will see an immediate leapfrog to over 10 times the aggregate data rate they had been experiencing before using the carrier’s cellular RAN/vRAN. In TDD spectrum, the performance increase is in both downlink and uplink. In FDD spectrum, the performance increase is in uplink spectrum only.

Yes, pCell vRAN works exceptionally well in multi-story structures, including tall buildings.
‍
pCell vRAN is inherently three-dimensional and distributing pWave RRUs vertically as well as horizontally results in enhanced performance (due to more angular diversity).
‍
In contrast, Cellular RANs/vRANs perform poorly in tall buildings because cellular networks are inherently two-dimensional in the horizontal plane. A phone in an upper story of a tall building in a city will have a near line-of-sight view to many cellular base stations, resulting in multi-cell interference and unstable handovers. Also, buildings with multi-floor DAS and small cells experience intercell interference among floors and, when installed in upper stories, among cells in the surrounding area.

No. pCell vRAN uses standard, non-synchronous Ethernet. While many cellular RANs and vRANs require IEEE 1588, SyncE, CPRI, or CPRI over Ethernet, pCell vRAN does not.

Conventional Ethernet switches and conventional fiber or copper cabling are used between the pCell vRAN servers and the PoE (Power Over Ethernet) switches that connect to the pWave RRUs (Remote Radio Units).

Ideally, the round-trip network latency (i.e., due to cable length and switching) between the pCell vRAN servers and the pWave RRUs should be < 100 µsec (< 0.1 msec). We’ve found this is generally achievable using on-premises Ethernet.

If higher latencies are unavoidable (e.g., if the pCell vRAN servers are off-premises or utilizing a high-latency communications transport), then pCell vRAN can compensate for higher latency with additional server computing resources and additional pWave RRUs. For example, performance with 5 msec of network latency is still quite usable for most use cases.

Independent of pCell vRAN latency requirements, ultra-low-latency Multi-Access Edge Computing (MEC) applications such as Extended/Augmented Reality require a very low-latency network between the pCell vRAN servers and the pWave RRUs.

Yes, pCell vRAN continues to work while user devices are in motion and/or when there is environmental motion.

pCell vRAN is constantly updating the waveforms that synthesize the pCells, calculating new waveforms in less than 1 millisecond (1 thousandth of a second) after receiving sounding signals from the user devices.

Yes. The pWave RRU components are available from -40°C to +85°C and pCell transmissions are impacted by inclement weather less than cellular technology in the same bands.

Given the current component shortage and initial customer needs, our inventory of pWaves currently only includes versions for indoor use or short-term outdoor use, e.g., outdoor music festivals. As components become available, we’ll be also building pWaves for permanent outdoor use.

pCell networks are dimensioned for the aggregate data rate required by the customer throughout the cover area, which is the data rate per pCell times the number of pCells.

The data rate per pCell is limited by the maximum data rate per LTE/5G user device. Most LTE/5G phones in use today have 1 uplink antenna and support up to 64 QAM (Quadrature Amplitude Modulation) downlink. To be compatible with all phones, pCell vRAN currently transmits 64 QAM to one antenna, resulting in a maximum of about 45-55 Mbps per user device in 20 MHz of spectrum, depending on LTE/5G protocol stack parameters.

Using 45 Mbps as the per-pCell data rate, a pCell vRAN with 32 concurrent pCells has an aggregate capacity of 45 Mbps x 32 = 1.44 Gbps in 20 MHz. Similarly, the aggregate capacity of a smaller pCell vRAN with 12 concurrent pCells would be 45 Mbps x 12 = 540 Mbps in 20 MHz.

The more pCells required by the customer, the more pWave RRUs and the more pCell vRAN server capacity is needed.

pCell vRAN multiplies system capacity by synthesizing multiple concurrent pCells and uses TDMA (Time Division Multiple Access) to share the number of concurrent pCells among the concurrent user devices.

For the purposes of an example, consider the unrealistic scenario where there are 32 phones, each running speed test at once (i.e., each demanding the maximum data rate at once) and there are 32 pCells, each delivering 45 Mbps in 20 MHz. In this situation, all 32 phones would each constantly be allocated pCell, and each would receive 45 Mbps at once, for an aggregate data rate of 45 x 32 = 1.44 Gbps.

Suppose now there are 64 phones, each running speed test at once. The pCell vRAN scheduler would use TDMA to cause each of the 32 pCells to jump back and forth between 2 phones, giving each phone a pCell 50% of the time, resulting in 45/2 = 22.5 Mbps per phone. Note that the aggregate data rate would be the same 22.5 x 64 = 1.44 Gbps because a pCell takes less than a nanosecond to jump from one phone to another, incurring no overhead for using TDMA.

In a real-world scenario, there are thousands of phones with each phone making widely varying different demands on the network, ranging from high data rate usage, such as streaming video to low data rate usage, such as receiving text messages. In this real-world scenario the pCell vRAN scheduler uses TDMA to allocate a pCell to each phone for the duration of time the phone needs based on its data demands. For example, a phone streaming 4K video would be allocated a pCell far more frequently than a phone receiving text messages. Effectively, the 32 pCells become a pool of resources that are allocated by the scheduler to meet demand. As in the previous examples, the aggregate data rate available for the thousands of phones is still 1.44 Gbps, and the vRAN scheduler uses TDMA to divide up this aggregate capacity among phones based on each one’s data demand.

Yes, the pCell vRAN scheduler can be configured to guarantee a highly reliable downlink and/or uplink data rate to specified user devices. Unlike cellular RANs/vRANs whose data rates vary widely through the coverage area, pCell vRAN delivers a highly uniform data rate throughout the coverage area, so user devices specified to receive a guaranteed downlink/uplink data rate will continue to reliably receive/transmit that data rate reliably throughout the coverage area.

For example, video cameras used as part of the live feed from a concert or other event can be given a guaranteed uplink data rate and can roam freely throughout the crowd with a reliable guaranteed data rate. Similarly, a vlogger in the crowd using their phone’s camera can be given a guaranteed uplink data rate. Security and emergency medical staff can be given guaranteed data rates to be sure they can always send and receive video. Transactional devices like credit card ticket readers can be given guaranteed data rates, even if they are small.

No. pCell vRANs have been tested in a wide range of indoor and outdoor scenarios with different types of antennas such as omnidirectional and patch antennas.

The choice of antennas to use in a pCell vRAN deployment depends on the deployment topology in order to achieve the desired coverage and capacity increase through the deliberate overlap and interference of wireless waveforms.

For instance the pCell vRAN deployment at SAP Center uses patch antennas affixed to the arena catwalk and pointing towards the seating area. We’ve tested omnidirectional antennas at SAP Center and they also work well, but they unnecessarily direct part of transmissions toward the ceiling, so patch antennas directed toward the seating area are more energy efficient

Future: pCell vRAN technology can improve Wi-Fi communications.

Future: pCell vRAN can dynamically track the absolute x, y, z position of user devices in real-time.

Future: pCell vRAN software can be ported to ARM CPUs, but Artemis is currently focused on using AMD and Intel x86 instruction set CPUs.

Future: pCell vRAN technology can improve LEO (Low Earth Orbital) satellite communications.

Future: pCell vRAN can transmit small amounts of power wirelessly to user devices.

Future: In theory, yes, but in practice GPUs are quite different architecturally than CPUs, so it would not be a direct port. Our initial focus with GPUs is to use them to support ultra-low-latency Multi-access Edge Computing (MEC) applications such as Extended/Augmented reality, much as GPUs are used today for Virtual Reality.