← HomeTR
GraduateFall 2026

Distributed Systems

Distributed Systems course

Language
Turkish (the paper may be written in Turkish or English)
Class meetings
Online: flexible working hours and online supervision
Contact
Questions by e-mail; announcements and submissions on this platform

Course description

Distributed systems are the foundation of today's computing infrastructure, from Netflix's video streaming and banking transaction systems to search engines and social media platforms. In this course each student investigates one sub-field in depth and produces an original piece of academic work.

The course is project-based and delivered fully online, with flexible working hours and online supervision. Students review the literature on a chosen topic, develop a research question grounded in a gap in that literature, design and carry out a suitable method, and submit a 25–30 page academic paper written in LaTeX at the end of the term. Along the way they learn to write a paper in LaTeX, review sources and present their findings effectively. The work is individual and develops technical knowledge, research skills and scientific writing together.

Research areas (suggestions; you may propose your own question):

  • Distributed databases: Data no longer fits on a single server; keeping and managing data across many machines. The CAP theorem, consistency models, sharding strategies, replication (Cassandra, MongoDB, CockroachDB, Google Spanner)
  • Distributed file systems: Storing petabytes of data reliably in the age of big data. Metadata management, fault tolerance, erasure coding, data replication (HDFS, Google File System, Amazon S3)
  • Content delivery networks (CDN): How is one video served to millions of users around the world at the same time? Edge servers, caching strategies, adaptive video streaming, load balancing (Akamai, Cloudflare, Amazon CloudFront)
  • Distributed machine learning: Billion-parameter models do not fit on a single GPU. Data, model and pipeline parallelism, federated learning (Ray, DeepSpeed, Horovod)
  • Cloud computing: Services at different levels of abstraction: IaaS, PaaS, SaaS, serverless, Kubernetes, autoscaling, multi-cloud and cost optimization
  • Edge computing: Moving computation from the cloud closer to the data source, for latency-critical applications. Edge–cloud hybrid architectures, resource allocation (IoT, autonomous vehicles, AR/VR)
  • Search engines: Distributed search architectures that answer billions of queries per second. Inverted indexes, ranking algorithms, real-time indexing (Elasticsearch, Solr)
  • Peer-to-peer systems: Systems that work without a central server. Distributed hash tables, consensus mechanisms, the free-rider problem (BitTorrent, blockchain, IPFS)
  • Internet of Things: Collecting, processing and analysing data from billions of sensors and devices. MQTT and CoAP, time-series databases, IoT security
  • Virtualization: Abstracting hardware resources. Hypervisors, containers, unikernels, virtual machine migration, network virtualization, container security
  • Smart cities: City-scale distributed systems for traffic, energy, the environment and public safety. Data integration, privacy, digital twins

When choosing a topic, consider both your interests and your existing knowledge.

A good research question is specific, answerable and not yet answered. Frame it concretely, for example “Does approach X improve metric Z in scenario Y?”. Examples: How do different consistency models affect performance? How can latency be reduced in geographically distributed systems?

Learning outcomes

On successful completion, students will be able to:

  1. 1Carry out a comprehensive and in-depth literature review in a chosen area of distributed systems
  2. 2Develop an original, clear and answerable research question grounded in a gap in the literature
  3. 3Design and carry out a method suited to the question (experimental, theoretical or survey)
  4. 4Analyse, visualize and consistently interpret the findings
  5. 5Write a 25–30 page academic paper in LaTeX in IEEE or ACM format, citing sources correctly
  6. 6Use AI tools as an aid, in line with the rules of academic integrity

Weekly schedule

WeekTopicReadings and work
1Choosing a topicChoose a topic that fits your interests and existing knowledge. Submit: 1–2 page summary and 5 academic papers
2–4Research and literature reviewRead and take notes, grow the reference list to 15; find a gap in the literature
5–6Method designExperimental (prototype, comparison), theoretical (modelling, analysis) or survey
7Interim report (midterm)8–10 pages: introduction and problem statement, literature review, research question, method. You receive feedback on this report
8–10Carrying out the studyWriting code, running experiments and simulations: prototype development, data collection and analysis, visualizing and interpreting results
11–12Writing the paperTitle and abstract, introduction, related work, proposed method, evaluation, discussion and conclusion. Use short, clear sentences
13Final checks and submission25–30 page LaTeX paper (PDF, figures, references); spelling, figure and table numbering, citations, reference format

Assessment

40%Midterm60%Final
Midterm100 points in total
  1. Interim report8–10 pages: introduction, literature, research question, method
    Assessed with 5 criteria
    • Breadth of the literature review20
    • Depth of the literature review20
    • Originality of the research question20
    • Clarity of the research question20
    • Suitability of the method20
    Submission 2 Oct – 20 Nov100 pts
Final100 points in total
  1. Final paper25–30 pages, LaTeX (IEEE/ACM)
    Assessed with 9 criteria
    • Breadth of the literature review10
    • Depth of the literature review10
    • Originality of the research question10
    • Clarity of the research question10
    • Suitability of the method10
    • Technical quality12,5
    • Reliability of results12,5
    • Clarity of presentation12,5
    • Overall coherence12,5
    100 pts

Resources

Course policies

Assessment: The interim report (midterm, 8–10 pages) and the final paper (25–30 pages) are graded with a rubric. Interim report criteria: breadth and depth of the literature review, originality and clarity of the research question, suitability of the method. The final paper is additionally assessed on technical quality (technical depth and correctness of the work), reliability of results (consistency and verifiability of the findings), clarity of presentation (writing quality and clarity of expression) and overall coherence (the integrity of the paper as a whole). Grade weights: interim report 40%, final paper 60%.

Academic integrity: Presenting someone else's ideas as your own, quoting without citation and copying another work are strictly forbidden. All work is checked with plagiarism detection software.

AI tools: Tools such as ChatGPT or Claude may be used as aids for developing ideas, drafting and language editing; their output must not be copied directly. Write the paper in your own words and with your own understanding.

Individual work: There is no group work. You may discuss with classmates, but each student's work must be their own.

Changing topic: Possible within the first two weeks, very difficult afterwards; decide early.

Coding is not required: A theoretical analysis or a comprehensive survey is also acceptable; what matters is an original contribution.

Language: The paper may be written in Turkish or English. Writing in English aligns with the international literature.

Submission format: The paper is prepared in LaTeX (Overleaf may be used; IEEE or ACM format recommended) and submitted as a PDF together with figures and references.