Erasure Coded Distributed Storage for Mobile Edge Networks

This direction is split into other directions: erasure-coded distributed storage. Its last version stays readable and citable here.

Version 1 of 169 papers83 conceptsfrozen 2026-10-04

In traditional wireless networks, data delivery is "connection-centric": a user maintains a link to a specific base station to download a specific file. However, as mobile data traffic explodes—the so-called "cellular spectrum crunch"—this model fails. Large-scale networks often suffer from "throughput collapse," where per-node capacity vanishes as the number of users grows.

To solve this, researchers like Alex Dimakis and Giuseppe Caire have proposed a shift toward "content-centric" architectures. Instead of chasing marginal gains in spectral efficiency, these systems exploit the massive growth in cheap storage by caching content at the network edge. The core mechanism enabling this is erasure coding, which transforms how data is stored and retrieved in dynamic, unpredictable environments.

The Intuition: From Specific Packets to "Any k" Fragments

The fundamental bottleneck in distributed caching is not just storage space, but the coordination required to ensure a user gets exactly what they need. In a simple replication scheme, a node either has the whole file or it doesn't. If a mobile user moves between cells, they must resume a specific download from a specific point.

Erasure coding, specifically Maximum Distance Separable (MDS) codes, changes the math of retrieval. As explained by Alexandre Graell i Amat and Amina Piemontese in MDS-Coded Distributed Caching for Low Delay Wireless Content Delivery, a file is partitioned into $k$ fragments and encoded into $n$ distinct symbols. The "magic" of MDS codes is that a receiver can reconstruct the original file by collecting any subset of $k$ symbols.

This property decouples specific packets from specific destinations. A user no longer needs "Packet 5"; they just need "any fragment I don't already have." This makes every encounter with a base station or another device productive, regardless of which specific pieces of the file they hold.

Harvesting Data Through Mobility

In a mobile ad-hoc network (MANET), node movement is usually seen as a challenge to be overcome. However, researchers like Konstantinos Poularakis and Leandros Tassiulas argue that mobility can be a powerful tool for Exploiting User Mobility for Wireless Content Delivery.

When files are erasure-coded and spread across the network, a moving user "harvests" fragments as they transition between different cells or encounter different peers. Shan Zhou and Lei Ying 0001 demonstrated that this "joint coding-scheduling" approach maximizes throughput under strict delay constraints. Because the user only needs a threshold number of fragments, the departure of a specific node or the loss of a specific link does not interrupt the retrieval process. This resilience allows the network to function as a robust distributed memory.

FemtoCaching: Storage at the Wireless Edge

The architecture for this approach often involves "helpers"—low-cost, small-cell base stations with high storage but limited backhaul. In the FemtoCaching framework, popular video content is distributed across these helpers using MDS codes.

This strategy solves two problems simultaneously:

  1. Backhaul Relief: By caching content closer to the user, the system reduces the load on the main cellular core.
  2. Convex Optimization: As Negin Golrezaei and colleagues noted, using coded fragments instead of whole files transforms the discrete, NP-hard problem of file placement into a continuous, solvable optimization problem. This allows for more efficient use of heterogeneous storage resources across the network.

Beyond Throughput: Privacy and Reliability

The benefits of coded distributed storage extend beyond speed. In Private Information Retrieval From a Cellular Network With Caching at the Edge, Eirik Rosnes and Siddhartha Kumar show that MDS codes can be used to hide a user's request from colluding edge nodes. By carefully placing coded fragments, the system can ensure that no single node (or small group of nodes) can determine which file the user is reconstructing.

Furthermore, this approach is being adapted for decentralized ledgers. Ravi Kiran Raman and Minghui Xu have proposed using erasure coding to scale blockchains. By allowing individual nodes to store only a fraction of the ledger while maintaining full recoverability, systems like EC-Chain can reduce storage requirements by over 90% without sacrificing security.

Conclusion

The shift toward erasure-coded distributed storage represents a move from "dumb" hardware to "smart" network-level logic. As Alex Dimakis famously argued, we can build reliable systems from "dumb," cheap sensors by using network coding to provide the necessary intelligence. By treating data as a "liquid" resource that can be reconstructed from any available fragments, these researchers have created a blueprint for wireless networks that are faster, more private, and inherently resilient to the chaos of mobility.

Papers in this version

Showing 60 of 69 papers.

History & citationVersion 1 · 2026-10-04

Split into other directions · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_erasure_coded_distributed_storage_for_mo_d767c0aa,
  title = {Erasure Coded Distributed Storage for Mobile Edge Networks},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/erasure-coded-distributed-storage-for-mobile-edge-networks/rdr_d767c0aa64264e1fbdedd5a2ae6cb639},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_erasure_coded_distributed_storage_for_mo_d767c0aa_v1,
  title = {Erasure Coded Distributed Storage for Mobile Edge Networks},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/erasure-coded-distributed-storage-for-mobile-edge-networks/rdr_d767c0aa64264e1fbdedd5a2ae6cb639/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}