We organized the opportunities into Research Strands (RS), which are lead by our team. Here, you can find excellent opportunities for your MSc, PhD and Post-doc careers, comprising all the areas of the SMARTNESS Center. For more details about each RS, please go to Research Strands.
DisCoNet: Distributed Computer Networking
Type: PhD
Title: Disaggregation and offloading of AR/XR applications to the programmable edge
Abstract: Disaggregation and offloading of parts of applications to the edge has been one of the key aspects for reducing latency and transfer of high quantities of data to the core. Applications such as the ones for AR/XR are very sensitive to latency and may benefit from running closer to the user mainly in a 5GB network. At the same time, we have witnessed a lot of new programmable hardware (switches, FPGAs, smartnics, IPUs) being deployed at the edge which may be used for running certain parts of the applications. This PhD project must then investigate the new set of programmable hardware and try to deploy part of the applications inside such hardware. There are some recent works that started to investigate this area and should be the starting point for the candidate.
Desired skills: Linux, virtualization, Python, computer networks. It is a must if the candidate has some knowledge on programmable hardware such as P4, Tofino and NetFPGA.
| Location | Funding | Starting date | Advisor | Contact | Monthly stipend (Brazilian Reais) |
| Sorocaba, SP | FAPESP | January/2023 | Fábio L. Verdi | verdi@ufscar.br | First year: R$ 3.462,60 Second to fourth years: R$ 4.285,50 Plus 20% annually for research contingency funds (Optionally) 1-Year Research Internship Abroad (BEPE) at a partner international institution |
Good references to start with:
- Towards a Fully Disaggregated and Programmable Data Center. APSys ’22: Proceedings of the 13th ACM SIGOPS Asia-Pacific Workshop on Systems.
- CLARA – Automated SmartNIC Offloading Insights for Network Functions. SOSP 2021 – Proceedings of the 28th ACM Symposium on Operating Systems Principles.
- Offloading distributed applications onto smartnics using ipipe. SIGCOMM 2019 – Proceedings of the 2019 Conference of the ACM Special Interest Group on Data Communication.
Type: Post-doc
Title: Functional decomposition of applications to programmable hardwares
Abstract: The candidate will work in the area of disaggregation of resources. Disaggregation is a concept where resources are separated into pools, facilitating elasticity, deployment and management. The main idea is to carry out the functional decomposition of applications into blocks of codes and map these blocks to different hardwares such as programmable network devices (e.g. Tofino switches), SmartNICs and FPGAs, which are used to perform network processing and other specialized functions, releasing the CPU to only handle applications. The focus of this research is to identify the blocks of a given application (to be defined), generate the Intermediate Representation (IR) using LLVM, and then compile the blocks to the right target.
Desired skills: Linux, virtualization, Python, computer networks. It is a must if the candidate has some knowledge on LLVM and Intermediate Representation, P4, Tofino and NetFPGA.
| Location | Funding | Starting date | Advisor | Contact | Monthly stipend (Brazilian Reais) |
| Sorocaba, SP | FAPESP | March/2023 | Fábio L. Verdi | verdi@ufscar.br | R$ 8,479.20 Plus 10% annually for research contingency funds |
Good references to start with:
- Towards a Fully Disaggregated and Programmable Data Center. APSys ’22: Proceedings of the 13th ACM SIGOPS Asia-Pacific Workshop on Systems.
- CLARA – Automated SmartNIC Offloading Insights for Network Functions. SOSP 2021 – Proceedings of the 28th ACM Symposium on Operating Systems Principles.
- Offloading distributed applications onto smartnics using ipipe. SIGCOMM 2019 – Proceedings of the 2019 Conference of the ACM Special Interest Group on Data Communication.
- E3: Energy-Efficient Microservices on SmartNIC-Accelerated Servers. Proceedings of the Usenix ATC 2019.
NEWTON: iN Edge netWork indusTrial automatioN
Type: PhD
Title: High-precision edge programmable dataplanes
Abstract: The proposed work plan builds upon the hypothesis that P4 datapaths can implement tasks to parse and extract the payload data, perform some calculations (e.g., distance, threshold, temperature), filtering, and much more, including crafting custom messages sent to the target endpoint (e.g. Robotic systems, IoT devices) or the controller, to minimize latency and deliver in-time and on-time network services. The high-precision hybrid edge programmable dataplane should deliver precise timing and latency of packet delivery to provide the required high-precision control loops. Expected advances include in-time and on-time network-provided packet delivery services for target throughput and acceptable packet loss, moving beyond the state of the art at IEEE TSN, IETF DETNET, and ITU-T future network design discussions. The resulting datapath architecture will be experimentally evaluated through prototype implementation in real commercial targets and 10G, 25G, and 100G workloads using high-precision traffic generators (e.g. OSNT, T-Rex) and real application traffic.
Desired skills: P4, Embedded systems, HW/SW co-design.
| Location | Funding | Starting date | Advisor | Contact | Monthly stipend (Brazilian Reais) |
| Campinas, SP | FAPESP | March/2023 | Christian Rothenberg | chesteve@unicamp.br | First year: R$ 3.462,60 Second to fourth years: R$ 4.285,50 Plus 20% annually for research contingency funds (Optionally) 1-Year Research Internship Abroad (BEPE) at a partner international institution |
Good references to start with:
- Network 2030 A Blueprint of Technology, Applications and Market Drivers Towards the Year 2030 and Beyond. 2019.
- Towards a Sweet Spot of Dataplane Programmability, Portability and Performance: On the Scalability of Multi-Architecture P4 Pipelines
C3LA: Cognitive Closed Control Loops Architecture for Edge IoV Services
Type: PhD
Title: Multi-Access Edge ML-based QoE Estimator
Abstract: The PhD student will investigate suitable methods and edge node component architecture capable of estimating the QoE as perceived per target users in access networks close to / served from a MEC platform where QoE estimation is carried. In addition to the engineering challenges of the MEC-based QoE estimator, multiple research questions are still open around the limits and applicability of ML to softwarized networks and end-to-end service lifecycle management. How to port techniques from one networking scenario, service, and time frame to another one in an effective manner is only one of the many-fold limitations in the state of the art on ML/AI for networking. Research questions on ML for CCL include QoE inference and prediction based on network-level data collection, as well as the ability to perform root cause analysis of service degradation, among many applications of ML for IoV
Desired skills: Machine Learning, Python, computer networks.
| Location | Funding | Starting date | Advisor | Contact | Monthly stipend (Brazilian Reais) |
| Campinas, SP | FAPESP | March/2023 | Christian Rothenberg | chesteve@unicamp.br | First year: R$ 3.462,60 Second to fourth years: R$ 4.285,50 Plus 20% annually for research contingency funds (Optionally) 1-Year Research Internship Abroad (BEPE) at a partner international institution |
Good references to start with:
- Intelligent Offloading in Multi-Access Edge Computing: A State-of-the-Art Review and Framework
- Machine Learning Approach to Estimate Video QoE of Encrypted DASH Traffic in 5G Networks
DEMIST: Distributed Edge Computing Swarm Intelligence
Type: PhD
Title: Distributed learning for vehicular networks
Abstract: This PhD student will focus on optimizing, running, and evaluating distributed learning mechanisms in vehicular network environments. The main objective is to investigate how vehicles can contribute to efficiently collect and process data to generate knowledge about many aspects in a smart city, such as traffic behavior and improvement through automated intersections, objects recognition and tracking, collision avoidance, pollution, and weather monitoring, and so on.
Place: Campinas, SP
Funding: FAPESP
Advisor: Luiz F. Bittencourt
Contact: bit@ic.unicamp.br
Good references to start with:
- Bonawitz et al. Towards Federated Learning at Scale: System Design. Proceedings of Machine Learning and Systems 1 (MLSys 2019).
- Tang et al. Comprehensive survey on machine learning in vehicular network: technology, applications and challenges. IEEE Communications Surveys & Tutorials.
Type: MSc
Title: Resource allocation policies for distributed learning for vehicular networks
Abstract: This MSc student will focus on designing resource allocation policies and network traffic optimization for distributed learning in vehicular networks. Data collected in vehicles may need to be shared with neighbors or with the cloud in order to improve decision-making aspects that involve more than one vehicle, as in traffic jam reduction and traffic flow optimization. Also, the distributed learning model should be updated from time to time, and update mechanisms for vehicular networks should also share data. Deciding when updates are needed will involve accuracy monitoring and models re-distribution along the edge and cloud.
| Location | Funding | Starting date | Advisor | Contact | Monthly stipend (Brazilian Reais) |
| Campinas, SP | FAPESP | March/2023 | Luiz F. Bittencourt | bit@ic.unicamp.br | First year: R$ 3.462,60 Second to fourth years: R$ 4.285,50 Plus 20% annually for research contingency funds (Optionally) 1-Year Research Internship Abroad (BEPE) at a partner international institution |
Good references to start with:
- Bonawitz et al. Towards Federated Learning at Scale: System Design. Proceedings of Machine Learning and Systems 1 (MLSys 2019).
- Tang et al. Comprehensive survey on machine learning in vehicular network: technology, applications and challenges. IEEE Communications Surveys & Tutorials.
ADVENTURE: Adaptive and Secure Network Applications over the Industrial Internet
Type: PhD
Title: Trustable devices communication and efficient cybersecurity mechanisms in 5G and B5G
Abstract: Future services in mobile networks will involve several partners to work (users, cloud providers, edge providers, highway concessionaires, etc.), and it is important to be sure that each node in the service is, indeed, the expected node. This will reduce the chances of security breaches and ensure the expected QoS level, considering the users’ identity. In this scenario, the Ph.D. student will propose mechanisms to allow the fast and secure identification of all the communication nodes. Innovative architectures and protocols such as Blockchain and SPDM (Security Protocol and Data Model) will be considered for the solutions. When providing URLLC and mMTC types of end-to-end services, mechanisms for protecting the network need to be much more efficient than the current ones to avoid excessive overheads that increase the latency, break the SLAs, or even cause a DoS in the system. The Ph.D. student will study the state-of-the-art cryptographic mechanisms and adapt them to future network architectures considering the levels of processing and latency expected in the network infrastructure.
Desired skills: Security, Linux, Computer Networks.
Place: São Paulo, SP
Funding: FAPESP
Advisor: Daniel Macêdo Batista
Contact: batista@ime.usp.br
| Location | Funding | Starting date | Advisor | Contact | Monthly stipend (Brazilian Reais) |
| São Paulo, SP | FAPESP | March/2023 | Daniel M. Batista | batista@ime.usp.br | First year: R$ 3.462,60 Second to fourth years: R$ 4.285,50 Plus 20% annually for research contingency funds (Optionally) 1-Year Research Internship Abroad (BEPE) at a partner international institution |
Good references to start with: