LOFAR Development Newsletter July 2025

Henrik Edler

Wide-field imaging with the international LBA stations is among the most ambitious plans for LOFAR2.0, with the vision of imaging substantial areas of the ultra-low frequency sky at sub-arcsecond resolution. In the past, the extreme ionospheric effects limited our ability to calibrate the international baselines to extremely bright sources (>10 Jy). However, recent progress in LOFAR calibration techniques showed that, already in before the LOFAR2.0 upgrade, calibration of fainter (≥0.5 Jy) sources can be achieved.”
At least in good ionospheric conditions, such sources are sufficiently common to calibrate a wide field of view. In a test using an 8h observation, a resolution of 0.9”x0.5” at a noise level of 650 µJy/beam was achieved between 42 and 66 MHz [see image]. This is the first-ever direction-dependent calibration of the international LBA stations. Together with the improvements the LOFAR2.0 upgrade will bring; it is an important step towards LBA VLBI as standard imaging mode in the LOFAR2.0 era.

Carla Baldovin

On June 2, DANTE held the Delta Critical Design Review. This was the follow-up review where the team could present the final tests of the AHBAFE based on the prototypes installed in 4 tiles in the LOFAR2.0 test station CS001. This milestone marked the end of development of the AHBAFE.
The review report states that “the design presented during CDR is very promising and ready for production”.
After closing the final actions, the team has now transferred the design to the next phase of producing several thousands of boards for the new stations in Italy and Bulgaria.

Carla Baldovin

The main achievements over the last couple of months are:

• The dismantling of LOFAR1 stations in the Netherlands is completed.

• Subrack integration for the Dutch stations completed.

• All stations in the Netherlands have now LOFAR2.0 subracks.

• The first batch of 10 local control units (LCU2) was delivered, set up and installed at the stations.

• Verification of CS002 started. The station was started up on June 24. This is the first one from the rollout campaign 2025. After performing integration tests, the station was handed over to the verification team. The team is taking time to carefully go over the full verification procedure, reducing the risks for the next stations.

Plans for the summer period:

• Continue the integration testing and verification of more LOFAR2.0 stations.

• As new LCU2s are received, they will be installed in new stations.

• Start installation of new power supplies that will be delivered in batches during the coming months.

• The new environmental controller (EC2) is in prototype phase; it will be tested in the laboratory and then in the LOFAR2.0 stations.

• Continue the integration of subracks for international stations.

• Organize the shipment of LOFAR2.0 hardware to Sweden. This will be the first delivery to our international partners.

Tom Kamphuis

Software Update – Steady Progress Across the Stack

Team Ruby continues to make significant strides toward the MPS. Over the past weeks, several important milestones have been reached:

Preprocessing Pipeline from TMSS

Change of pipeline frameworks to run the preprocessing pipeline again using the new pipeline language (Common Workflow Language / CWL). This is important for CEP6 and running new pipelines later.

Rolling Out New LCU Hardware

LCU installation in the stations has started. These upgraded units provide the computational backbone for LOFAR stations and are essential for running the enhanced LOFAR2.0 software stack.

StationTest Now Integrated

We have integrated StationTest, our internal validation suite, into the main software environment. This enables better verification of the stations, helping us to spot issues early on.

TMSS FAAI Authentication

FAAI implementation to allow users to use their university account to access LOFAR services, now also connected to TMSS.

Arno Schoenmakers

The implementation of the firmware and OPC-UA interface for the Transient Buffer functionality for LIFT is on schedule. By the end of June, we expect to have all functionality working, including sending the transient buffer data over the data ring between the FPGAs on a single Uniboard, to a data switch. This will then be working in a laboratory environment. What remains is integration into real LOFAR2.0 stations, which will be done in August and September.

Ilse van Bemmel

Multicast roll-out

A key aspect of the LOFAR2.0 design is the high-speed network that enables an order of magnitude increase in data transfer rates. As part of the CEP Project the core and concentrator node switches have been replaced with hardware that meets the new requirements. In addition, the network will use Multicast. A sender only needs to transmit the data once, and interested parties can subscribe to the Multicast group. The same technology is used for TV broadcasts and within Internet Service Provider networks for IPTV.

On 20 May we had a dedicated busy day for the roll-out of Multicast in the LOFAR2.0 network. This change affects all sub-systems of LOFAR and is therefore a tedious task. The team spent the full day upgrading the different components and fixing resulting problems. An additional layer of security was added to reduce the risk of failures and external attacks.

Enabling Multicast in the full system resulted in many small and some larger hurdles which took another day to resolve, but on Thursday 22 May the LOFAR2.0 system was up and running again. This change has achieved an important milestone for LOFAR2.0 development and allows new LOFAR2.0 stations to be connected directly to the new Multicast network during roll-out.

Figure 1 shows how data streams from the router receiving the station data into the Cobalt correlator nodes (green and blue lines in the image). Center top is the router, which talks to the top-rack switch ACX-B3-0. From the switch the Cobalt2 nodes are receiving multiple data streams with different data rates. Node CBT206 is subscribed to 6 and 9 Gbps, at the same time Cobalt node CBT201 subscribes to 3 and 6 Gbps. Receiving these two simultaneous streams is the trademark of a Multicast system.

Hardware roll-out for CEP6 and system services

A batch order was placed for the CEP6 and Telescope Operation (TO) services hardware at the end of April. This included a partial (~40%) CEP6 cluster, which despite its small size is already as powerful as CEP4. Having a smaller system will enable proper benchmarking in an operational environment to finetune the specifications for a follow-up hardware order later this year. Over time this will set the basis for a system that can be incrementally upgraded.

The hardware arrived at the end of May (Figure 2), allowing us to start roll-out with convenient extra time before the summer period. After installing and connecting everything at CIT in Groningen, the hardware is currently undergoing acceptance testing and should be ready for software roll-out by the end of June. Software provisioning will go in order of priority: first the TO hardware, which includes hypervisors for the nomad, consul and minio systems, then the TO S3-storage, and finally the CEP6 compute and storage. The work should result in a system ready for transition of operations from CEP4 to CEP6.

The transition from CEP4 to CEP6 is planned in weekly phases during September, as new hardware and new software will result in new issues to resolve. The first week the CEP6 systems will be switched on in parallel to CEP4. After that each Tuesday we will only use CEP6, identify the main problems, use the subsequent week to fix these. This should gradually move operations from CEP4 to CEP6, resulting in the shutdown of CEP4 by the end of September.

Operations on CEP6 will look different as user access will be more strictly controlled. For the operational CEP6 compute cluster no user access will be allowed, an experimental CEP6 compute environment will be made available for use cases not supported by standard operations. To support commissioning the current CEP6 hardware will be set up as experimental, even so, access will be limited to active commissioners and will be more tightly managed than before.

Arno Schoenmakers

The project management plan (PMP) for the QuickBlitz/AARTFAAC2.0 project has been finished and is under review by the ASTRON MT. This includes a first more detailed planning for the delivery of this instrument (see Figure).

The first step to be taken is to create a more detailed system description which we call the AARTFAAC2 MPS, or Minimal Production System. We will use the MPS to set development priorities for the development team. Boudewijn Hut, the LOFAR2.0 system engineer, and his system engineering team will help us in this process to ensure a seamless integration in the LOFAR2.0 design and functionality.

Also, we will decide on the correlator hardware over the summer, so we can start up the purchasing process. We will be advised by John Romein, who will implement the correlator functionality on this system later in the project.